Shape of the thecal sac: L3/4 interspace compared with L4/5*

Abstract

We retrospectively reviewed 60 normal magnetic resonance imaging scans to assess the anatomical shape of the thecal sac at the L3 ⁄ 4 and L4 ⁄ 5 vertebral interspaces. In all cases the thecal sac was oval at L3 ⁄ 4 but in 26 (43%; 95% CI 31-55%) the thecal sac changed from oval at the L3 ⁄ 4 interspace to triangular at L4 ⁄ 5 (with the apex of the triangle presenting to the posterior epidural space). We propose that this anatomical variant would make it more difficult to obtain cerebrospinal fluid at the lower level, as a slightly lateral approach could lead to identification of the epidural space but failure to puncture the thecal sac. This may offer an explanation for a 'dry tap' when a lower interspace is chosen. Damage to the conus is a potentially serious complication of inserting a needle into the thecal sac. This can occur during spinal or combined spinalepidural (CSE) anaesthesia or lumbar puncture. Following a number of case reports [1,2] it was recommended by Reynolds that these procedures should be performed at the lowest possible palpable interspace, which could be between the spinous processes of L4 and L5 or occasionally lower. It has been found that, especially during the needle-throughneedle CSE technique, the epidural space can sometimes be located but cerebrospinal fluid (CSF) not obtained because of failure to puncture the dura as the spinal needle protrudes from the Tuohy needle. The incidence of this is reported to be between 5% and 16% [3][4][5]. We have observed that this problem may occur more frequently when the lowest palpable interspace is used, with easier thecal puncture when the procedure is performed one interspace higher. We therefore undertook a retrospective review of normal lumbar magnetic resonance imaging (MRI) scans to determine if there was an anatomical reason underlying these observations. MethodsApproval was obtained from the Trust's Audit Department and the Local Research Ethics Committee considered formal review unnecessary. Magnetic resonance imaging scans of the lumbar spine with T1 and T2 weighted sagittal and axial images were retrospectively reviewed. All patients had been referred for MRI scan for the investigation of lower back pain and the scans had been reported as anatomically normal. An initial number of 100 patients was proposed but as our search criteria were very strict, looking only for normal MRI studies performed for non-neurological conditions, a total of 60 scans was found after hand searching several years of reports. A hypothetical calculation, based on our findings after 60, showed that doubling our sample size further would reduce our 95% CI by only ± 3%.On an initial review of a small number of scans it was noted that in some cases there was a change in shape of the spinal canal and thecal sac within it, on the axial images, from oval at the mid space of L3 ⁄ 4 to triangular at the mid space of L4 ⁄ 5. The review of subsequent scans was then based on this initial observation and a series of

Ultrasonography helps emergency physician identify the best lumbar puncture site under the conus medullaris

“…Interestingly, the conus medullaris ends in most cases in the intervertebral space L1-L2 but in 2 to 5% of cases it descends to below the vertebral body of L2 [9]. Furthermore, the shape of thecal sac varies in lower interspace making the success of the lumbar puncture more uncertain [10]. We thus believe that the main advantage of ultrasound marking is related to the identification of lower interspinous level than those traditionally chosen after anatomical marking.…”