Importance of Surgical Order for Minimizing Vascular Injury During the L5-S1 Approach in Multilevel Oblique Lateral Interbody Fusion Surgery

Abstract

Objective

In oblique lateral interbody fusion (OLIF) surgery at the L5-S1 level (OLIF51), anatomical complexity and the possibility of vascular injury during retraction of the common iliac vein (CIV) make the surgery challenging. We radiologically evaluated patients who underwent OLIF surgery to determine approaches that can make OLIF51 surgery easier during multilevel OLIF.

Methods

We retrospectively analyzed 275 consecutive patients who underwent OLIF surgery between September 2014 and December 2019. The distance between the left and right CIVs (dCIV) was measured using an axial image at the L5 lower endplate level, and the height of the iliocaval junction (hCIV) was measured from the L5 lower endplate to the iliocaval junction in the sagittal image. The sum of anterior disc height of each level (sADH) was calculated.

Results

Eighty-two patients (33 males and 49 females) were enrolled. The number of three- (L2-3-4-5), two- (L3-4-5), and one-level (L4-5) fusions was 13, 21, and 48, respectively. Changes between the pre- and postoperative sADH, dCIV, and hCIV values were 17.1±4.7, 7.7±3.5, and 13.1±4.7 mm in three-level fusion; 10.6±4.1, 5.6±3.7, and 7.0±3.1 in two-level fusion; and 4.3±2.5, 3.3±2.7, and 3.0±2.0 mm in one-level fusion, respectively. As the number of surgical levels increased, the changes in sADH, dCIV, and hCIV significantly increased.

Conclusions

The dCIV and hCIV values increased when the upper segment underwent surgery before OLIF51 during multilevel OLIF.

INTRODUCTION

Oblique lateral interbody fusion (OLIF) is a minimally invasive surgical technique that can fuse the L2 to S1 vertebrae in a single position.¹³⁾ Many studies have reported the clinical and radiological results of L2-5 OLIF (OLIF25) and L5-S1 OLIF (OLIF51), and they are known to have excellent outcomes in correcting lumbar lordosis.⁸,¹⁰,¹¹,¹³⁾

In recent years, with an increasingly aging society, the use of short- and long-level fusion surgery is increasing. In addition, the importance of lumbar lordosis correction for short- and long-level fusion surgery is increasing. In particular, lordosis correction of the lower lumbar region (L4-5-S1) is the most important part.¹,⁵,¹⁴⁾ In this regard, OLIF surgery is advantageous in correcting the lower lumbar angle. However, especially in the case of OLIF51, the possibility of vascular injury during retraction of the common iliac vein (CIV) as well as the anatomical complexity make it difficult for the operator to attempt OLIF51.¹⁰⁾ Several studies have reported preoperative radiological evaluation to reduce vascular injury during OLIF51 surgery²,³⁾; however, no study has evaluated the OLIF51 approach by assessing vascular shape changes during multilevel spinal fusion surgery.

We radiologically evaluated patients who underwent OLIF surgery to determine how to simplify OLIF51 surgery during multilevel OLIF surgery.

MATERIALS AND METHODS

Study participants

This study was approved by the Institutional Review Board of Chung-Ang University Hospital (2101-001-19347). We retrospectively analyzed 275 consecutive patients who underwent OLIF surgery performed by two surgeons (SWP and MJK) between September 2014 and December 2019 in Chung-Ang university hospital. Patients with acute traumatic fracture, spinal metastasis, infection, and a history of previous lumbar fusion surgery were excluded. Considering the possibility of postoperative anatomical change due to CIV retraction during the L5-S1 approach, patients who underwent OLIF51 were also excluded.

Surgical technique

The surgical technique for OLIF25 has been published previously.⁶⁾ Briefly, patients were positioned in the right lateral decubitus position under general anesthesia. Using C-arm imaging, a 4–5 cm oblique skin incision was made from the anterior margin of the target disc. The external oblique, internal oblique, and transverse abdominis muscles were split along directions of muscle fibers. The target disc was approached anteriorly through the retroperitoneal space, and a cage (Clydesdale; Medtronic, Minneapolis, MN, USA) packed with a demineralized bone matrix (Grafton; Medtronic, Minneapolis, MN, USA) was inserted using an orthogonal maneuver after the disc was removed. In case of multilevel surgery, the operation was performed from the lowest disc and proceeded upward by the same way as mentioned herein.

Radiologic evaluation for anterior space at the L5-S1 disc level

The distance between the left and right CIVs (dCIV) and the height of the iliocaval junction (hCIV), which is known to be related to vascular injury, were measured.²⁾ The dCIV was measured using an axial image at the L5 lower endplate level, and the hCIV was measured from the L5 lower endplate to the iliocaval junction in the sagittal image (FIGURE 1). The anterior disc height (ADH) is the distance between the lower endplate of the upper vertebra and the upper endplate of the lower vertebra of the operated segment; the ADH per segment (sum of anterior disc heights of each level; sADH) was summed.

FIGURE 1. Measurements of dCIV and hCIV.

dCIV and hCIV were measured on the preoperative and postoperative non-enhance spine CT. CIA (white solid line circle) could be easily identified because it looks round, and the oval-shaped CIV (white dotted line circle) is observed behind the CIA. The distance between the left and right CIVs was measured using an axial image at the L5 lower endplate level and designated as dCIV (2-way arrow) (A). The hCIV (2-way arrow) was measured from the L5 lower endplate to the iliocaval junction in the sagittal image by using cross reference line (white dotted line) (B).

dCIV: distance between the right and left common iliac vein, hCIV: height of the iliocaval junction, CT: computed tomography, CIA: common iliac artery.

All radiologic factors were measured from computed tomography (CT) images taken pre- and postoperatively. The measurements were undertaken by three independent examiners (2 spine surgeons and 1 neurosurgical resident) twice at 1-week intervals.

Statistical analysis

Pre- and postoperative radiological values were compared using the Wilcoxon signed-rank test in each fusion level and the Kruskal-Wallis test among fusion levels. Cage height and angle were compared using the Kruskal-Wallis test. Post hoc testing was conducted using the Mann-Whitney U test for factors that were found to be significant in the Kruskal-Wallis test. For all analyses, p<0.05 was considered statistically significant. The intraclass correlation coefficient (ICC) values were graded using previously described semiquantitative criteria (0.90–1.0: excellent, 0.70–0.89: good, 0.50–0.0.69: fair/moderate, 0.25–0.49: low, and 0.0–0.24: poor).

RESULTS

The average age of the 82 (33 male and 49 female) enrolled patients was 65.5±7.5 (range: 45–77) years. The number of total fusion levels was 129, and the primary diagnosis of each fusion level was central canal stenosis in 59 (45.7%), foraminal stenosis in 33 (25.6%), degenerative spondylolisthesis in 23 (17.8%), and intervertebral disc herniation in 14 (10.9%). The number of three- (L2-3-4-5), two- (L3-4-5), and one-level (L4-5) fusions were 13, 21, and 48, respectively.

The mean height and angle of the used cages were 13.4±0.8 mm and 10.6°±1.3° in three-level fusion, 13.2±0.8 mm and 10.9°±1.4° in two-level fusion, and 13.5±1.1 mm and 11.9°±2.3° in one-level fusion, respectively, and the differences were not significant (p=0.156 and p=0.064, respectively; TABLE 1).

TABLE 1. Patient characteristics.

Characteristics		Values	p-values
Number of patients		82	-
Age (yr)		65.5±7.5	-
Sex ratio (male:female)		33:49	-
Number of total fusion level		129	-
	3 levels fusion (L2-3-4-5)	13
	2 levels fusion (L3-4-5)	21
	1 level fusion (L4-5)	48
Diagnosis of each level			-
	Central canal stenosis	59 (45.7%)
	Foraminal stenosis	33 (25.6%)
	Degenerative spondylolisthesis	23 (17.8%)
	Intervertebral disc herniation	14 (10.9%)
Cage height (mm)			0.156
	3 levels fusion	13.4±0.8
	2 levels fusion	13.2±0.8
	1 level fusion	13.5±1.1
Cage angle (°)			0.064
	3 levels fusion	10.6±1.3
	2 levels fusion	10.9±1.4
	1 level fusion	11.9±2.3

Changes in ADH, dCIV, and hCIV

Preoperative sADHs were 24.5±5.6, 17.2±5.2, and 9.9±2.7 mm in three-, two-, and one-level fusions, respectively, which were significantly increased postoperatively to 41.7±2.5 (p=0.001), 27.8±2.4 (p<0.001), and 14.2±1.6 mm (p<0.001), respectively.

The preoperative dCIVs were 33.3±9.3, 29.1±10.2, and 29.2±10.6 mm in three-, two-, and one-level fusions, respectively. The postoperative dCIVs were significantly increased to 41.0±9.1 (p=0.001), 34.6±9.0 (p<0.001), and 32.5±10.0 mm (p<0.001) in three-, two-, and one-level fusions, respectively.

The preoperative hCIVs were 36.6±10.9, 36.9±9.7, and 38.5±11.8 mm in three-, two-, and one-level fusions, respectively, which were significantly postoperatively increased to 49.7±11.4 (p=0.001), 43.9±9.8 (p<0.001), and 41.5±11.7 (p<0.001), respectively, in three-, two-, and one-level fusions (TABLE 2).

TABLE 2. Pre- and postoperative radiologic measurements in each fusion level.

Variables		Preop	Postop	p-value
Three-level fusion
	sADH (mm)	24.5±5.6	41.7±2.5	0.001
	dCIV (mm)	33.3±9.3	41.0±9.1	0.001
	hCIV (mm)	36.6±10.9	49.7±11.4	0.001
Two-level fusion
	sADH (mm)	17.2±5.2	27.8±2.4	<0.001
	dCIV (mm)	29.1±10.2	34.6±9.0	<0.001
	hCIV (mm)	36.9±9.7	43.9±9.8	<0.001
One-level fusion
	sADH (mm)	9.9±2.7	14.2±1.6	<0.001
	dCIV (mm)	29.2±10.6	32.5±10.0	<0.001
	hCIV (mm)	38.5±11.8	41.5±11.7	<0.001

Preop: preoperative value, Postop: postoperative value, sADH: sum of anterior disc heights of each level, dCIV: distance between right and left common iliac vein, hCIV: height of the iliocaval junction.

The intra-observer and inter-observer ICCs were 0.85–0.92 and 0.78–0.86, respectively (TABLE 3).

TABLE 3. Strength of agreement of inter- and intra-observer analysis for parameters.

Variables	Intra-observer ICC	Inter-observer ICC
ADH	0.92	0.86
dCIV	0.85	0.82
hCIV	0.87	0.78

ICC: intraclass correlation coefficient, ADH: anterior disc height, dCIV: distance between right and left common iliac vein, hCIV: height of the iliocaval junction.

Changes between the pre- and postoperative sADH, dCIV, and hCIV values were 17.1±4.7, 7.7±3.5, and 13.1±4.7 mm for three-level fusion, 10.6±4.1, 5.6±3.7, and 7.0±3.1 for two-level fusion, and 4.3±2.5, 3.3±2.7, and 3.0±2.0 mm for one-level fusion, respectively (FIGURE 2). As the number of surgical levels increased, the changes in the sADH, dCIV, and hCIV also significantly increased (p<0.001; FIGURE 3).

FIGURE 2. Pre- and postoperative changes on CT images.

In one-level fusion surgery, dCIV was increased from 6.54 mm preoperatively (A) to 12.96 mm postoperatively (B). In two-levels fusion surgery, hCIV was increased from 21.3 mm preoperatively (C) to 33.4 mm postoperatively (D).

CT: computed tomography, dCIV: distance between the right and left common iliac vein, hCIV: height of the iliocaval junction, Preop: preoperative, Postop: postoperative.

FIGURE 3. Changes (Δ, postoperative value-preoperative value) in sADH, dCIV, hCIV.

As the number of surgical segments increased, the left and right CIV distances widened significantly, and the height of CIV bifurcation increased.

sADH: sum of anterior disc heights of each level, dCIV: distance between the right and left common iliac vein, hCIV: height of the iliocaval junction, CIV: common iliac vein.

DISCUSSION

In multilevel spinal fusion surgery, correction of lumbar lordosis is an important factor related to patients’ postoperative quality of life; in particular, the importance of lower lumbar lordosis correction is emphasized.⁵,¹⁴⁾ OLIF25 and OLIF51 are widely used minimally invasive spinal fusion techniques that are reportedly superior to conventional surgical techniques for correcting lumbar lordosis.⁴,⁸,¹¹⁾

OLIF enables lumbar fusion surgery from the L2 to the S1 vertebrae in a single position, and OLIF51 is excellent for lordosis correction of the L5-S1 segments.⁸,⁹,¹¹⁾ However, the risk of vascular injury, which may occur while performing OLIF51, is high due to the need to retract the left CIV while approaching the disc space and the anatomical complexity.³,¹³⁾

The incidence of CIV injury during anterolateral surgical approaches was found to be 8.6%–11.5%, and risk factors for CIV injury included a low iliocaval junction, narrow CIV bifurcation, insufficient perivascular fat tissue, and vascular anomaly.²,³⁾

Capellades et al.²⁾ classified patients into the following 12 groups: based on hCIV (very high, high, low, and very low) and based on dCIV (medial, intermediate, and lateral). The 5 groups of them (high/medial, low/medial, very low/medial, very low/intermediate, and very low/lateral) were reported that there was a risk of vascular injury due to the required CIV manipulation.²⁾ On using this subgroup classification in our study dataset, we found that there were 3 (23.1%), 6 (28.6%), and 11 (22.9%) patients with preoperative vascular injury risks in the three-, two-, and one-level fusion groups, respectively. There was a significant postoperative reduction in the number of patients with vascular injury risks from 3 to 1 (7.7%, p=0.035), from 6 to 3 (14.3%, p=0.023), and from 11 to 8 (16.7%, p<0.001) patients in the three-, two-, and one-level fusion groups, respectively (FIGURE 4).

FIGURE 4. Changes in vascular injury risk of the L5-S1 approach according to Capellades et al.²⁾ .

According to Capellades et al.,²⁾ the 5 subgroups marked in gray (high/medial, low/medial, very low/medial, very low/intermediate, and very low/lateral) had risks of vascular injury because CIV manipulation was required. When our study results were subjected to the above subgroup classification, the risk of vascular injury was significantly reduced after all types of fusion surgery.

The illustration is newly drawn and attached.

CIV: common iliac vein, Preop: preoperative, Postop: postoperative, N: number.

Chung et al.³⁾ reported that the probability of vascular injury was increased when the amount of perivascular fat tissue was low. In this study, since perivascular fat tissue was not measured, it was impossible to directly confirm whether the increment of the postoperative dCIV and hCIV was associated with the amount of perivascular fat tissue. However, we did find that the dCIV and hCIV increased in all patients who underwent OLIF25; this suggests that considering the degree of CIV adhesion related to the amount of perivascular fat tissue is not necessary in patients with degenerative spinal disorders.

The results of previous studies have shown that a relatively taller cage could be inserted using lateral approaches, including OLIF, compared with that using posterior approaches, thus increasing the postoperative disc height significantly.⁷,¹²⁾ In our study, as in previous studies, the sADH increased with the surgical level, and it was confirmed that the dCIV and hCIV also increased with the sADH. When OLIF was performed at the L4-5 level, the dCIV and hCIV increased by 11.3% and 8.1%, respectively, as compared with their preoperative values. When OLIF was performed at the L3-4-5 level, the dCIV and hCIV increased by 19.2% and 19.0%, respectively, from their preoperative values. When OLIF was performed at the L2-3-4-5 level, the dCIV and hCIV increased by 23.1% and 35.8%, respectively, from their preoperative values. Therefore, it is expected that the risk of CIV injury during OLIF51 could be reduced if OLIF is performed on the upper segments and finally on the L5-S1.

Since the CT images were taken with the patient in the supine position, the vascular structure may appear different in the right lateral decubitus position, which is the patient position during the OLIF surgery. CT images taken with the patient in the right lateral decubitus position could better reflect the actual vascular structure during OLIF. However, since our study is retrospective in nature and since most CT scans are performed with the patient in the supine position in clinical settings, it may be more reasonable to establish a surgical plan based on the CT images taken in the supine position.

The present study is retrospective study with relatively small volume. Therefore, a larger-prospective multicenter study is necessary to provide more verified information. Additionally, it is expected that the results of our study will be more clearly verified if the intraoperative CT scan is used to investigate changes of CIV during surgery.

CONCLUSION

The dCIV and hCIV increased if surgery was performed on the upper segment first before OLIF51 was undertaken during multilevel OLIF surgery. It is expected that it will be helpful if these points were considered when planning the surgery.