It is commonly accepted that the epidural space is a continuous compartment containing fat, lymphatics, arteries, loose areolar connective tissue, spinal nerve roots, and an extensive plexus of veins [2, 3]. However, recent studies have suggested that meningo-vertebral ligaments are present between the lumbosacral spinal canal wall and the surrounding dura. Meningo-vertebral ligaments scattered in the epidural space with adipose tissue on their surface separate the epidural space into compartments of different sizes and shapes [4], which results in discontinuity of the epidural space, resulting in uneven distribution of injectate, thereby contributing to unilateral or partial anaesthesia after injection of local anaesthetic [5]. The failure rate for neuraxial labor analgesia is reported as high as 12 % [6]. When the initial epidural placement and infusion does not result in an adequate block, anesthesiologists often perform 2 different maneuvers to salvage the catheter: partial withdrawal of the catheter or infusion of additional medication [7]. If all these 2 maneuvers couldn’t lead to a successful analgesia, anaesthetist should try other maneuvers such as replacement of the epidural catheter or a combined spinal epidural anaesthesia.
Meningo-vertebral ligaments located within the epidural space may contribute to epidural catheter placement failure [8]. The location of these structures can be unpredictable with 38 % of meningo-vertebral ligaments located in the middle of the subdural cavity, 32 % in the left side subdural cavity, and another 30 % next to right side of subdural cavity [9]. Our study found three epidural catheters curled into a circle in 3 cases, with one catheter curled into the shape of a “9”. We hypothesise that these structures may contribute to catheter knotting within the epidural space [10]. The conclusion of some reports is that insertion of excessive amounts of catheter into the epidural space is a causative factor in knot formation [11]. Some authors have recommended the insertion of no more than 4 cm of catheter into the epidural space and some others no more than 5 cm [12, 13]. According to the study of Aslanidis T et al. [14], the 6 cm of catheter left into the epidural space in the current case was not excessive. This may not become apparent until removal of the catheter, at which time, catheter fracture may occur. Irregardless, when removing an epidural catheter, one should hesitate when resistance is encountered, change position and then slowly attempt to remove the catheter again. If necessary, surgical assistance should be sought [15, 16].
Epidural catheter was placed at the L3-L4 interspace. As it was seen at Table 1, the highest frequency of catheter curling occured at the same level. Bridenbaugh et al. reported that 48 % of the catheters coiled at the insertion site [17]. One possible reason is that a large number of meningo-vertebral ligaments at the L3-L4 interspace blocked the catheter tip. At the same time, an abrupt change of direction of the catheter tip in the epidural needle tip make it difficult to insert the catheter. Maybe, all the two reasons resulted in a higher incidence of catheter curling at the L3-L4 interspace.
Others have reported that 5 % of epidural catheters travel caudally or exit the intervertebral foramen [18]. In our study, there were two cases in which the epidural catheter traveled into the intervertebral foramen, and we hypothesise that this course may have been due to variant meningo-vertebral ligaments. In such cases, the epidural injectate would course laterally into the intervertebral foramen resulting in nerve root rather than epidural spread of the injectate.
In this study, we observed five cases where the catheter punctured the internal vertebral venous plexus. Unintended venous injury may occur during lumbar epidural catheter placement, with an estimated incidence rate of 5–7 % [6]. Yaakov berlin et al. found that patients whose catheters were threaded 7 cm into the epidural space had the highest incidence (17.5 %) of intravenous catheter placement as compared to patients whose catheters were threaded 3 or 5 cm [18]. They speculated that more catheter was threaded into the epidural space the greater was the likelihood of encountering a vein. If not recognized, local anaesthetics will enter into the systemic circulation, potentially resulting in seizure, cardiovascular toxicity, circulatory failure or death [19, 20].
Our study found that the meningo-vertebral ligaments directly connected to the wall of the spinal microvasculature in the location of lamina and ligamentum flavum. These structures were entwined with, but not adherent to the vascular walls. Earlier studies have reported that small blood vessels were found in fetal tissue specimens of meningo-vertebral ligaments [21, 22]. When performing epidural catheter placement, if the needle inadvertently damages blood vessels connected to this ligament, the risk of developing an epidural hematoma is hypothesised to increase. Vascular structures connecting to the meningo-vertebral ligaments are mostly venous, these veins lack valves and exist in a space where compression can result in catastrophic outcomes.
No inadvertent dural puncture was reported in our study. The incidence of inadvertent dural puncture ranges between 0.19-0.5 % of epidural catheter placements. Adverse events may result from direct mechanical injury or adverse physiological responses. Inadvertent dural puncture and postdural puncture headache, direct neural injury, total spinal anesthesia, and subdural block have been commonly reported.
Meningo-vertebral ligaments become thicker as they occur more caudally, with the thickest ligament at L5-S1. At the caudal levels, the meningo-vertebral ligaments are firmly connected with the ligamentum flavum, which may be the reason that intraoperative dural tearing occurs more commonly during surgery at the L4-L5 and L5-S1 levels [1].
The major limitation of this study is the lack of clinical data from live patients. Preserved cadaveric specimens, while the only suitable study population for the research which we undertook, are limited by the changes in the tissue structure with death and fixation of the body with formalin. It is well known that tissue fixation and its associated desiccation can lead to tissue shrinkage. The use of formalin may have contributed to an overestimation of the strength of the ligaments than they are in vivo. Maybe, it is one of the reasons why the incidence (10 %) of intravenous catheter placement in our study was much lower than that (17.5 %) in vivo.