If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Corresponding author. Department of Orthopedics, Tongji Hospital, Tongji medical College, Huazhong University of Science and Technology, No.1095 Jie Fang Avenue, Wuhan, China. Tel: +86-027-83665219.
The anterior approach at L5-S1 has many advantages, however, vascular complications are challenging for spinal surgeons who may not be familiar with the variability of vascular anatomy. There are three different anterior approaches (intra-bifurcation approach and extra-bifurcation: left-, and right-sided prepsoas approaches) described in previous studies to respond to the variability of anterior vascular anatomy for reduction in vascular injury, while no guidance for the choice of approach preoperatively.
PURPOSE
To analyze the anatomical feasibility of three anterior approaches to access the L5-S1 disk space according to a practical framework.
STUDY DESIGN
Retrospective study.
PATIENT SAMPLE
Lumbar magnetic resonance imaging (MRI) from patients who visited our outpatient clinic were reviewed, with 150 cases meeting the inclusion criteria.
OUTCOME MEASURES
The following radiographic parameters were measured on axial T2-weighted MRI at the lower endplate of L5 and the upper endplate of S1: width of the vascular corridor, position of the left and right common iliac vein (CIV), and presence of perivascular adipose tissue (PAT). Moreover, we designed a safe line to evaluate the feasibility of left- and right-sided prepsoas approaches. Cases of lumbosacral transitional vertebrae were identified.
METHODS
The feasibility of the intra-bifurcation approach was determined by the width of the vascular corridor, presence of PAT, and the position of the CIV. The feasibility of the prepsoas approach was determined by the relative position of the CIV to the safe line, presence of PAT, and the intersection point of the CIV and vertebral body.
RESULTS
Sixty-eight percent, 64.7%, and 75.3% cases allowed the intra-bifurcation, left-, and right-sided prepsoas approach to L5-S1, respectively. The cases in this study had at least one of three anterior approaches to access L5-S1 disk space, and 74% of cases had more than one anatomical feasibility of anterior approach. The right-sided prepsoas approach was feasible in the majority of cases because of the vertical course of the right CIV with a significantly higher proportion of presence of PAT. Patients with lumbosacral transitional vertebrae (24 cases) may prefer the prepsoas approaches, and only six cases (25.0%) were determined to be feasible for the intra-bifurcation approach.
CONCLUSIONS
Our study proposes a practical framework to determine whether the three different anterior approaches are feasible access at L5-S1. According to the framework, all cases had the anatomical feasibility of using an anterior approach to access L5-S1, and three-fourths of cases had a replaceable anterior approach when encountering intraoperative difficulties.
Anterior lumbar interbody fusion (ALIF) at L5-S1 level has gained in popularity because it allows direct exposure of the disk space, offers superior correction of sagittal and coronal alignment, has high fusion rates, and avoids the risk of direct neural injury and posterior muscle damage [
]. However, potential risk of vascular injury remains challenge for spinal surgeons. Vascular injury in ALIF is reported to range from 1.5% to 15.6%, and the majority of vascular injuries are caused by laceration of the vessel during disk exposure [
]. An adequate exposure of the disk space between the bifurcation of the great vessel is necessary for the safe introduction of the interbody cage. However, a previous anatomical MRI study reported that one-third of the patients had a low or very low iliocaval junction position with a narrower vascular window [
], and posterolateral mobilization of the common iliac vein (CIV) is inevitable with an increase in vascular injury.
To minimize mobilization of the CIV in patients with vascular structures obstructing the anterior surgical space, oblique lumbar interbody fusion (OLIF) through the lateral surgical corridor external to the left CIV in L5-S1 was described by mobilizing the left CIV to the midline of disk space [
], simultaneously avoiding superior hypogastric plexus injury. Moreover, Tannoury et al. reported that either the left- or right-sided lateral pre-psoas surgical corridor can be used in minimally invasive anterior to the psoas (MIS-ATP) fusion at L5-S1 with no major vascular injury encountered [
]. Berry et al. recommended OLIF through the right-sided pre-psoas approach to access L5-S1 because the right CIV has a vertical course and is more predictable [
]. However, the choice of the three approaches in previous studies was based on the authors’ personal experience, and there is no guidance for spinal surgeons to choose an approach that avoids vascular injury during disk exposure.
Thus, this study aimed to provide a practical framework for spinal surgeons to guide the choice of safe approach based on analyzing the anatomical feasibility of avoiding vascular injury in the three anterior approaches (intra-bifurcation and extra-bifurcation: left-, and right-sided pre-psoas approaches) according to the practical framework.
Materials and methods
Patient selection
This retrospective observational imaging study obtained the ethical approval from the Ethics Committee of our hospital (TJ-IRB20210321). Lumbar spine magnetic resonance imaging (MRI) of consecutive patients who visited our outpatient clinic between January 2021 and June 2022 were reviewed.
The inclusion criteria were as follows: 1) age ≥18 years, 2) back pain and/or leg pain >1 month, 3) L5-S1 spondylolisthesis (grade Ⅰ or Ⅱ), 4) both lumbar MRI films were clear and axial T2-weighted MRI scans at the L5-S1 level a field of view containing the iliac wing on the left and right sides, 5) no previous lumbar surgery, 6) no lumbar scoliosis, and 7) no other obvious issues that can alter the results of the study.
Demographic data such as age and sex were obtained.
Radiographic measurement
For lumbar MRI, we obtained two images of axial T2-weighted MRI scans at the L5-S1 level (L5, lower endplate of L5, and S1, upper endplate of S1). The following data were obtained for each patient from the two images. The presence of perivascular adipose tissue (PAT) is defined as a fat plane visible between the vertebral body and CIV. We proposed the safe line to assess the feasibility of left or right-side pre-psoas approaches. The safe line is defined as a line cross the left or right highest point of iliac wing and the intersection point of “facet line” [
] and posterior border of vertebral body. Based on the safe line, CIV was defined as above, cross, and below the safe line. All measurement methods are shown in Fig.1.Width of vascular corridor, defined as the distance from the left CIV to the right CIV.
1.
Distance from the midline to the left CIV
2.
Distance from the midline to the right CIV
3.
Distance from the vertebral body to the left and right CIV in patients with the presence of PAT
4.
Distance from the left and right CIV to the posterior vertebral body line
5.
The width of cage corridor, defined as the vertical distance from the edge of the vertebral body to the intersection point of “facet line” and posterior border of vertebral body
6.
Tilt angles of the left and right safe line
Fig.1The measurement of radiographic parameters. 1) Width of vascular corridor; 2) Distance from the midline to the left common iliac vein (CIV); 3) Distance from the midline to the right CIV; 4) Distance from the vertebral body to the left and right CIV in patients with the presence of perivascular adipose tissue (PAT); 5) Distance from the left and right CIV to posterior vertebral body line; 6) The width of cage corridor; 7) Tilt angles of the left and right safe line.
] were defined as lumbosacral transitional vertebrae (LSTVs).
All radiographic data were measured using the MicroDicom viewer (Version 3.4.7) by two trained spine surgeons with years of clinical work experience, and the radiographic data were measured twice individually, with an interval of 2 weeks. A consensus was reached after discussion in cases of disagreement regarding the presence of PAT.
We proposed a practical framework for spinal surgeons to guide the choice of approach preoperatively (Table1). The subjects in this study were categorized into three groups (intra-bifurcation, left-, and right-sided pre-psoas groups) according to the feasibility of anterior approaches.
Table 1Framework to guide the choices of anterior approach to access the L5-S1 disk space
Intra-bifurcation approach
Left-sided pre-psoas approach
Right-sided pre-psoas approach
1. The width of vascular corridor ≥35mm
1. The left CIV above the safe line
1. The right CIV above the safe line
2. The width of vascular corridor ≥25mm
2. The left CIV cross the safe line
2. The right CIV cross the safe line
3. The presence of perivascular adipose tissue at left or right side
3. The presence of perivascular adipose tissue at the left-side and the central point of the left CIV above the safe line
3. The presence of perivascular adipose tissue at the right side and the central point of the right CIV is above the safe line
4. The left CIV is not beyond midline
4. The intersection point of the left CIV and vertebral body is above the safe line
4. The intersection point of the right CIV and vertebral body is above the safe line
Meet condition 1, or 2+3+4
Meet condition 1, or 2+3, or 2+4
Meet condition 1, or 2+3, or 2+4
Cases meeting the conditions in the axial images of lumbar MRI at both the lower endplate of L5 and the upper endplate of S1 were considered to be anatomical feasible for this approach.
All statistical analyses were performed using the SPSS version 21 (IBM Corp. Armonk, NY, USA). Intra-observer and inter-observer reliability were assessed using the intraclass correlation coefficient (ICC) for numeric variables and kappa statistics for the presence of PAT. Continuous variables are shown as mean±standard deviation, and normality tests was performed using Kolmogorov-Smirnov test. The paired t-test and chi-square test or Fisher's exact test were used to compare the variables between the left and right side in the two axial images. The independent sample t-test and the chi-square test or Fisher's exact test were used to compare the variables between patients with and without LSTVs. Statistical significance was set at p<.05.
Results
A total of 233 lumbar MRI was collected in this study, and 83 patients were excluded due to age <18 years (2 patients), previous lumbar surgery (28 patients), lumbar scoliosis (26 patients), lack of a field of view containing the iliac wing on the left or right sides (24 patients), and ambiguous of MRI films (3 patients). Eventually, 150 cases were included in this study (82 women and 68 men; mean age: 51.2±15.9 years; range, 18–84 years old). The intra-observer and inter-observer reliability of the two spine surgeons are shown in Table 2. A high level of consistency was observed, with the intra-observer and inter-observer reliability ranging from 0.782 to 0.992.
Table 2Inter-observer and intra-observer reliability of the two observers
Inter-observer
Intra-observer (Observer A)
Intra-observer (Observer B)
Parameters
ICC
ICC
ICC
L5
Width of vascular corridor
0.935
0.970
0.992
Left
Distance from midline to the CIV
0.937
0.977
0.980
Distance from CIV to the posterior vertebral body line
0.936
0.970
0.973
The width between the vertebral body and the CIV in cases with PAT
0.945
0.949
0.917
The tilt angle of the safe line
0.886
0.945
0.959
The width of cage corridor
0.783
0.951
0.886
Right
Distance from midline to the CIV
0.826
0.916
0.954
Distance from CIV to the posterior vertebral body line
0.941
0.970
0.968
The width between the vertebral body and the CIV in cases with PAT
0.960
0.972
0.906
The tilt angle of the safe line
0.882
0.921
0.994
The width of cage corridor
0.782
0.949
0.861
S1
Width of vascular corridor
0.895
0.975
0.993
Left
Distance from midline to the CIV
0.909
0.979
0.984
Distance from CIV to the posterior vertebral body line
0.889
0.978
0.949
The width between the vertebral body and the CIV in cases with PAT
0.984
0.972
0.971
The tilt angle of the safe line
0.910
0.938
0.960
The width of cage corridor
0.811
0.943
0.898
Right
Distance from midline to the CIV
0.834
0.916
0.961
Distance from CIV to the posterior vertebral body line
0.903
0.961
0.963
The width between the vertebral body and the CIV in cases with PAT
0.955
0.893
0.917
The tilt angle of the safe line
0.917
0.912
0.943
The width of cage corridor
0.798
0.921
0.918
Kappa values
Kappa values
Kappa values
L5
Left
The presence of PAT
0.827
0.853
0.879
Right
The presence of PAT
0.861
0.929
0.935
S1
Left
The presence of PAT
0.850
0.882
0.835
Right
The presence of PAT
0.869
0.956
0.878
ICC, intraclass correlation coefficient; CIV, common iliac vein; PAT, perivascular adipose tissue.
Vascular corridor and the position of the left, and right CIV
The average width of vascular corridor was 30.1 mm at the L5, and 37.2 mm at the S1. The left CIV was located at 10.3 mm and 17.0 mm from the midline at L5 and S1, respectively. The left CIV was often flat compared the right CIV at both L5 and S1 (74.0% vs. 25.3%, p<.01; 39.3% vs. 26.0%, p=.019), and the left CIV was significantly close to the midline than the right CIV (19.7 mm and 20.2 mm), as shown in Table 3. The distance from the left and right CIV to the posterior vertebral body line were not significant different at L5, however, the left CIV was located more posteriorly than the right CIV at S1.
Table 3The measurement of variables at the lower endplate of L5 and the upper endplate of S1
The presence of PAT and the width between the vertebral body and the CIV
50.0% and 27.3% of the patients did not have PAT on the left side of L5 and S1, respectively, and the proportion of cases absence PAT on the left side was significantly higher than that on the right side (50.0% vs. 16.0%, p<.01; 27.3% vs. 8.0%, p<.01). In the cases with the presence of PAT, the width between the vertebral body and the CIV was significantly higher on the right side of L5 and S1 than those on the left-side (2.5 vs. 1.2 mm, p<.01; 3.1 vs. 2.3 mm, p<.01, respectively).
Anterior approaches to access L5-S1
According to the framework, 68.0%, 64.7%, and 75.3% of cases have the anatomical feasibility of intra-bifurcation, left-, and right-sided pre-psoas approach to access L5-S1, respectively. The cases in this study had at least one of three feasible anterior approaches. 26.0%, 40.0%, and 34.0% of cases had one, two, and three anterior approaches to L5-S1, respectively (Fig.2A). In the cases with one anterior approach to access L5-S1, 61.5% of cases were intra-bifurcation approach, and 12.8%, and 25.6% of cases were left or right-sided pre-psoas approaches, respectively (Fig.2B). In the cases with two approaches, 55% of cases were left and right-sided pre-psoas approaches, 13.3% were intra-bifurcation and left-sided pre-psoas approaches, and 31.7% were intra-bifurcation and right-sided pre-psoas approaches (Fig.2C).
Fig.2The feasibility of anterior approaches to access L5-S1 disk space.
There were 24 (16.0%) cases were determined LSTVs with a significant narrower of vascular corridor comparing cases without LSTVs at both L5 and S1(21.3 vs. 31.7 mm, p<.01; 28.3 vs. 38.9 mm, p<.01, respectively). The incidence of cases without PAT on the left side of the S1 was significantly higher than that on the right side (62.5% vs. 20.6%, p<.01). Additionally, cases with LSTVs may prefer the pre-psoas approaches that only six of cases (25.0%) were determined to be feasible access using the intra-bifurcation approach to L5-S1 (Table 4).
Table 4Comparison of variables between cases with and without the LSTVs
Although technique improvement has led to a growing interest in the application of anterior approaches at L5-S1, vascular injury, especially injury of the CIV, remains a major concern of the anterior approaches. Understanding the variability in the vascular anatomy of the lumbosacral spine is helpful for spine surgeons to avoid catastrophic complications.
Standard lumbar spine MRI can provide information on the position and morphology of the vascular anatomy, which has been widely used in previous anatomical studies of the lumbosacral spine [
]. However, most of anatomical studies have been based on axial imaging taken at the midpoint of the L5-S1 disk space. The disk space exposure ranged from the lower endplate of L5 to the upper endplate of S1, and the axial image at the midpoint of the L5-S1 disk space provided a limited representation of the position and morphology of the iliocaval vasculature. In this study, we chose axial images at the lower endplate of L5 and upper endplate of S1 to evaluate the operative window. We believe that understanding the anatomy of the border of operative window is helpful for preoperative evaluation of the iliocaval vasculature.
The average width of the vascular corridor was 30.1 mm at L5 and 37.2 mm at S1. It was similar to the study by Nagamatsu et al. that 29.7 mm at L5 and 36.9 mm at S1 in the non-enhanced CT–MRI medical image [
]. In addition, we describe the distribution of the left and right CIV from the lower endplate of L5 to the upper endplate of S1, as shown in Fig.3. The left CIV is often flat and located closer to the midline than the right CIV at both L5 and S1. From L5 to S1, the left CIV moved laterally and posteriorly. However, the right CIV was more vertical with less movement. The presence of PAT indicates that adipose tissue surrounds the CIV, which can provide the safely retracted space during disk exposure and prevent CIV injury [
]. In this study, we found that the proportion of presence of PAT was higher at S1 than at L5, and higher on the right side than that on the left side (Table 3). It was also reported by the study of Song et al. that 50.7% of patients had PAT on the left side, which was significantly lower than that on the right side (88.3%) [
]. Thus, the discrepancy in the proportion of presence of PAT between the left and right sides and between the cranial and caudal sides should be considered preoperatively.
Fig.3Distribution of the left and right CIV at the lower endplate of L5 and upper endplate of S1.
There are three different anterior approach techniques to respond to complex and variable anterior vascular anatomy, and each approach has shown relative advantages. The vascular complications and neurological deficits, stratified by the anterior approaches used in previous studies, are listed in Table 5 [
Single position lateral decubitus anterior lumbar interbody fusion (ALIF) and posterior fusion reduces complications and improves perioperative outcomes compared with traditional anterior-posterior lumbar fusion.
]. However, the choice of approach in previous studies was based on the author's personal experience, neither providing guidance nor analyzing the feasibility of the three approaches. A practical framework to guide the choice of approach is helpful to avoid encountering intraoperative difficulties that require modification of the preoperative surgical planning [
]. We proposed a practical framework for spinal surgeons to guide the choice of safe approach based on analyzing the anatomical feasibility of avoiding vascular injury in the three anterior approaches. In this framework, the width of vascular corridor, position of the CIV, and presence of PAT were the key factors in choice of the intra-bifurcation approach. Although the use of the intra-bifurcation approach in the lateral decubitus position has been increasing in recent years, an adequately wider vascular corridor(≥25 mm) is still needed [
Single position lateral decubitus anterior lumbar interbody fusion (ALIF) and posterior fusion reduces complications and improves perioperative outcomes compared with traditional anterior-posterior lumbar fusion.
Anterior column reconstruction of the lumbar spine in the lateral decubitus position: anatomical and patient-related considerations for ALIF, anterior-to-psoas, and transpsoas LLIF approaches.
]. The absence of PAT and the CIV crossing the midline preclude safe intra-bifurcation access L5-S1 because it is challenging for spinal surgeons to safely mobilize the CIV outside the operative window [
Single position lateral decubitus anterior lumbar interbody fusion (ALIF) and posterior fusion reduces complications and improves perioperative outcomes compared with traditional anterior-posterior lumbar fusion.
Fig.4A: Cases with a wider vascular corridor was feasible access L5-S1 disk space through the intra- bifurcation approach. B: Cases with a narrower vascular corridor, absence of PAT, and the CIV cross the midline, were difficult to preform intra-bifurcation approach at L5-S1.
Additionally, we designed a safe line to assess whether pre-psoas approaches were feasible. The safe line is similar to a preset trajectory of the interbody cage. Based on our clinical practice, the iliac wing rather than the large bulky of psoas muscles, may disturb the insertion of cage in pre-psoas approaches. A higher tilt angle of the safe line induced a more oblique trajectory with a posteriorly located cage (Fig 5). A posteriorly located cage will undermine the correction of sagittal alignment and increase the risk of cage subsidence [
]. When the cases have a CIV above the safe line, the pre-psoas approach can be easily accessed. By contrast, in cases with the CIV blows the safe line, the pre-psoas approach may not be feasible. For cases with a safe line crossing the CIV, two points should be considered to avoid excessive mobilization of the CIV: the central point of the CIV should be above the safe line in cases with PAT, and the intersection point of the CIV and vertebral body should be above the safe line in cases without PAT (Fig 6).
Fig.5Case 1: A 72 years old female underwent L3-S1 fusion due to disc herniation with instability. The L5-S1 OLIF was performed using the left-sided pre-psoas approach. A higher safe line tilt angle (38.9°) was observed on the postoperative lumbar MRI with a posteriorly located cage. Case 2: A 55 years old female underwent L5-S1 OLIF using the left-sided pre-psoas approach due to lumbar spondylolysis. A lower safe line tilt angle (26.7°) was observed on the postoperative lumbar MRI with an anteriorly located cage.
Fig. 6For cases with a safe line crossing the CIV, the central point of the CIV (A) and the intersection point of the CIV and vertebral body (B) should be evaluated to determine the feasibility of the pre-psoas approach.
In this study, we found that all cases had at least one of the three anterior approaches to access the L5-S1 disk space according to the framework, and 40.0%, and 34.0% of cases had two and three approaches, respectively. This suggests that all cases had the feasibility of using an anterior approach to access L5-S1, and 74.0% of cases can perform an alternative anterior approach to access L5-S1 when intraoperative difficulties are encountered, leading to abortion of the preoperative surgical plan. Berry et al. reported one patient encountered intraoperative difficulty when attempting OLIF via left-sided intra-bifurcation technique because it was difficult to retract the left CIV, and using the left-sided pre-psoas technique was eventually successful [
]. Among the three approaches, the right-sided pre-psoas approach was feasible access for most of cases because of the vertical course of the right CIV with a significantly higher proportion of presence of PAT. Although a more medial position of the left CIV can increase the feasibility of the left-sided pre-psoas approach, the absence of PAT and a more posterior position at S1 limits the application of the left-sided pre-psoas technique. The intra-bifurcation approach was feasible in 68.0% of cases, and the right side may be a better choice because the lateral position of the right vessel and 77.7% most medially positioned vessels on the right side was arteries [
]. Moreover, a cadaver study by Paraskevas et al. found that the superior hypogastric plexus mainly located on the left-side of the midline and suggested that the operation should be performed along the right CIV [
The feasibility of anterior approach in patients with LSTVs is variable in patients without LSTVs. The location of the iliocava junction is significantly lower in patients with LSTVs [
]. Chung et al. reported that the left CIV was potentially difficult to mobilize in 74.2% of patients with LSTVs, whereas it was no requirement or easy to mobilize in 81.1% of patients without LSTVs. Moreover, 93.5% of patients with LSTVs underwent the OLIF through the lateral window at L5-S1 and 83.8% of patients without LSTVs through the central window [
]. In this study, only 25.0% of patients with LSTVs were determined to be feasible for intra-bifurcation approach to L5-S1 according to the framework. Thus, the pre-psoas approach may be a better choice for patients with LSTVs.
The present study had two main limitations. First, the configuration of the CIV in the lateral decubitus position is different from that in the supine position [
Movement of abdominal structures on magnetic resonance imaging during positioning changes related to lateral lumbar spine surgery: a morphometric study.
]. Simulating lumbar MRI in the lateral decubitus position can more directly demonstrate the position and morphology of the CIV, however, this is challenging in a clinical setting. Second, patients’ demographic data such as age, height, and body mass index may be useful to assess the technical difficulty of accessing the L5-S1 disk space, but we did not assess the variables in this study. Although a large sample size was used in this study, a prospective study with a larger sample size should be performed to confirm the validity of the framework.
Conclusion
Our retrospective study investigated the anterior vascular anatomy of the lumbosacral spine and proposed a practical framework for spine surgeons to guide the choice of approach (intra-bifurcation, left-, and right-sided pre-psoas approaches) to access the L5-S1 disk space. According to the framework, all cases had the anatomical feasibility of using an anterior approach to access L5-S1 disk space, and three of the four cases had a replaceable anterior approach when encountering intraoperative difficulties. Among the three approaches, the right-sided pre-psoas approach was feasible in the majority of cases (75.3%). Patients with LSTVs were preferred access to L5-S1 using the pre-psoas approach.
Acknowledgment
No funds were received to support this study.
References
Brau SA.
Mini-open approach to the spine for anterior lumbar interbody fusion.
Single position lateral decubitus anterior lumbar interbody fusion (ALIF) and posterior fusion reduces complications and improves perioperative outcomes compared with traditional anterior-posterior lumbar fusion.
Anterior column reconstruction of the lumbar spine in the lateral decubitus position: anatomical and patient-related considerations for ALIF, anterior-to-psoas, and transpsoas LLIF approaches.
Movement of abdominal structures on magnetic resonance imaging during positioning changes related to lateral lumbar spine surgery: a morphometric study.
Bold values indicate statistical significance p<.05.
00075-X&id=gr1.jpg){kind=link}
00075-X&id=gr2.jpg){kind=link}
00075-X&id=gr3.jpg){kind=link}
00075-X&id=gr4.jpg){kind=link}
00075-X&id=gr5.jpg){kind=link}
00075-X&id=gr6.jpg){kind=link}