Introduction

The lower lumbar spine is a major site of pathology in spine disorders, and surgery is frequently performed in this area. Anterior spinal interbody fusion is an effective treatment method for degenerative spine diseases and spinal deformities [1, 2], and recent methodological and technical innovations in spine surgery have rendered the anterior spinal approach safe. In particular, the lateral approach to anterior spinal fusion through the psoas muscle, extreme lateral interbody fusion (XLIF) [3], is becoming widely performed.

Intervertebral discs are composed of avascular tissue; they are safely and easily accessible and are, therefore, considered reasonable sites for spinal fusion. They can be approached from anteriorly, laterally, transforaminally, or posteriorly. Vascular injury can be an unexpected complication during the transpsoas approach to accessing the intervertebral disc laterally, [48] but because the lumbar artery, rather than the large vessels, may be a more frequent site of bleeding during the transpsoas approach, it should be understood in detail. However, there are few anatomic studies of the morphologic relationship between the branches of the lumbar artery and the intervertebral disc. In this study, we investigated the morphology of the branches of the lumbar artery at the lateral sides of the intervertebral disc and its clinical importance.

Materials and methods

We studied 88 sites (86 lumbar arteries; 2 sites demonstrated absence of the artery) at the third and fourth lumbar vertebrae bilaterally in 22 formalin-fixed cadavers. Specimens with transitional vertebrae were excluded. To identify each lumbar artery from the vertebra through the vertebral foramen, lumbar spines were divided into right and left at the midline of the vertebral bodies and placed in the lateral position. The third and fourth lumbar arteries at the L3 and L4 levels were exposed by carefully removing the psoas muscle and fatty tissues. The branches of the lumbar artery crossing the lateral sides of the intervertebral discs, including muscular branches, anastomotic branches, and branches supplying the spinal nerve and plexus (BSNP), were visualized. The iliolumbar artery running upward on the L4–5 intervertebral disc was also identified. The capillary branches and muscular branches extending into the psoas muscle shortly after the bifurcation were excluded. The specimens were evaluated for the presence of branches coursing vertically over the intervertebral discs.

Results

Muscular branches having a lumen structure longer than 2 cm crossed vertically over the middle third of the intervertebral disc in 3 of 88 sites (3.4 %) (Fig. 1). Anastomotic branches flowing into the caudal adjacent foramen or connecting with the iliolumbar arteries ran downward along the posterior third of the intervertebral disc in 13 of 88 sites (14.8 %,) (Fig. 2a). Iliolumbar arteries ran upward along the L4 spinal nerve on the L4–5 intervertebral disc in 2 of 88 sites (2.3 %) (Fig. 2b).

Fig. 1
figure 1

Photograph showing the muscular branch coursing vertically over the middle of the L4–5 intervertebral disc

Fig. 2
figure 2

Photographs showing a anastomotic branch and b iliolumbar artery coursing vertically over the posterior third of the intervertebral disc

A BSNP was seen in 25 of 44 sites (56.8 %) at L3 and in 40 of 44 sites (90.9 %) at L4, or 65 of 88 (73.9 % overall). The BSNPs originated at the lumbar artery and then ran downward, either through the posterior third of the intervertebral disc before the anterior edge of the spinal nerve, or at an extraforaminal site behind the anterior edge of the spinal nerve. The former pattern was found in 18 of 65 branches (27.7 %) (Fig. 3a) and the latter in 47 of 65 branches (72.3 %) (Figs. 3b, 4a). According to these results, the branches of the arteries crossed vertically on the posterior third of the intervertebral disc in 8 of 44 sites (18.2 %) at L3 and 18 of 44 sites (40.9 %) at L4, or 26 of 88 sites (29.5 % overall) (Table 1; Fig. 4b).

Fig. 3
figure 3

BSNPs originating from the lumbar artery. Difference between sites in which the BSNP courses a along the posterior third of the intervertebral disc and b extraforaminally behind the anterior edge of the spinal nerve. BSNP, branch supplying the spinal nerve and plexus

Fig. 4
figure 4

Schemata depicting a BSNPs and their locations running vertically in front of or behind the anterior edge of the spinal nerve; and b branches of the lumbar artery running down the middle of the disc and the posterior third of the disc laterally. BSNP branch supplying the spinal nerve and plexus

Table 1 Summary of arterial branches coursing vertically over the intervertebral discs of the lower lumbar spine

Discussion

XLIF is a novel, minimally invasive method in spinal fusion surgery for which good biomechanical effects have been reported. However, an incidence of vascular complications of 0.10 % was revealed in a survey of 13,004 patients whose procedures were performed by an experienced surgeon [8]. A few studies have demonstrated a safe zone for avoiding injuries of the retroperitoneal large vessels, which can be identified using preoperative magnetic resonance imaging (MRI) to assess the relative positions of the adjacent neurovascular structures with respect to the each intervertebral disc level [9, 10]. Although injury to large vessels must always be prevented, injury to the lumbar artery occurs more frequently than injury to large vessels during the transpsoas approach [5]. Furthermore, few reports have demonstrated the presence and precise morphology of the branches of the lumbar artery at the extraforaminal area at the disc level [11, 12]. The results of the present study clearly demonstrate the existence of branches of the lumbar artery running along the discs.

Although the anatomy and blood supply of the sciatic nerve have been described in detail [13], there have been no detailed anatomic studies of the branches of the lumbar artery and blood supply to the lumbar plexus. Arslan et al. observed that the anastomotic arteries ran vertically approximately 4 mm anterior to the base of the transverse process [12]. In an angiographic study of lumbar artery anastomoses, Ratcliffe et al. demonstrated that this anastomosis exists at all levels but is most pronounced between the arteries at L4 and L5, suggesting that the risk of bleeding was higher during surgery involving L4 and L5 than at other levels [14]. Tezuka et al. showed various vascular patterns around the lower lumbar spine in their anatomical analysis using contrast-enhanced computed tomography (CT) and identified several variations of distributing arteries from the lumbar artery to the posterior element of L4–5 [15]. Our study also revealed that the branches of the lumbar artery appeared more frequently on the posterior third of the L4–5 intervertebral disc than at the L3–4 level.

Our study also suggests that care should be taken when approaching disc through the psoas muscle and performing contralateral disc excision through the disc space because intraoperative bleeding and postoperative retroperitoneal hematoma caused by lumbar artery injury have been reported [5, 6]. Further, posterior disc penetration, posterior cage placement, and oblique cage insertion can be risk factors not only for neurological complications due to spinal nerve injuries [16, 17], but also for vascular complications. These studies suggest the importance of preoperative MRI or contrast-enhanced computed tomography to detect abnormal arteries and that it is essential to prepare for unexpected bleeding from the branches of the lumbar arteries during spine surgery via the transpsoas approach. The findings of the present anatomic study bring to spine surgeons an awareness of the existence of the branches of the lumbar artery coursing vertically over the intervertebral discs, which is important when performing spinal fusion surgeries via the transpsoas approach.

This study has several limitations. First, this anatomical study is based on gross observation of branches that have a lumen structure. Microstructural arteries must exist with greater frequency than that shown in this analysis. Second, the study sample was small, and it is possible that there are other variations in the morphology of the branches of the lumbar artery.