International consensus for a dissection room quality system (DRQS): A Delphi panel study.

Dissection Rooms (DRs) are key facilities that allow teaching and research on human anatomy, where students and researchers work with human bodies to acquire, increase, or create new knowledge. Usually, DRs work with a Body Donation Program (BDP), where living donors bequeath their bodies for use in teaching and research after they expire. Despite DRs being part of universities worldwide, no common guidelines, regulations, or quality management systems (QMS) exist that could be applied to different countries. With that purpose in mind, we aimed to develop a QMS that could be applied to DRs globally, using a Delphi panel to achieve consensus about the items that should constitute the QMS. The panel was constituted by 20 anatomy professors from 20 different countries, and the 167 standards to create the rules or guidelines that constitute the QMS were divided in five categories: direction, body donation, students, instructors, and research. After two rounds of revisions, 150 standards were considered "essential" or "important" by more than 70% of the participants, thus being incorporated to the Dissection Room Quality System (DRQS). The results of this panel represent a minimum list of items of the DRQS for improving the functioning of DRs globally.

Midbrain and pons MRI shape analysis and its clinical and CSF correlates in degenerative parkinsonisms: a pilot study.

OBJECTIVES: To conduct brainstem MRI shape analysis across neurodegenerative parkinsonisms and control subjects (CS), along with its association with clinical and cerebrospinal fluid (CSF) correlates. METHODOLOGY: We collected demographic and clinical variables, performed planimetric and shape MRI analyses, and determined CSF neurofilament-light chain (NfL) levels in 84 participants: 11 CS, 12 with Parkinson's disease (PD), 26 with multiple system atrophy (MSA), 21 with progressive supranuclear palsy (PSP), and 14 with corticobasal degeneration (CBD). RESULTS: MSA featured the most extensive and significant brainstem shape narrowing (that is, atrophy), mostly in the pons. CBD presented local atrophy in several small areas in the pons and midbrain compared to PD and CS. PSP presented local atrophy in small areas in the posterior and upper midbrain as well as the rostral pons compared to MSA. Our findings of planimetric MRI measurements and CSF NfL levels replicated those from previous literature. Brainstem shape atrophy correlated with worse motor state in all parkinsonisms and with higher NfL levels in MSA, PSP, and PD. CONCLUSION: Atypical parkinsonisms present different brainstem shape patterns which correlate with clinical severity and neuronal degeneration. In MSA, shape analysis could be further explored as a potential diagnostic biomarker. By contrast, shape analysis appears to have a rather limited discriminant value in PSP. KEY POINTS: • Atypical parkinsonisms present different brainstem shape patterns. • Shape patterns correlate with clinical severity and neuronal degeneration. • In MSA, shape analysis could be further explored as a potential diagnostic biomarker.

Endoscopic endonasal surgery to treat intrinsic brainstem lesions: correlation between anatomy and surgery.

OBJECTIVE: The endoscopic endonasal approach (EEA) has been proposed as an alternative in the surgical removal of ventral brainstem lesions. However, the feasibility and limitations of this approach to treat such pathologies are still poorly understood. This study aimed to report our experience in five consecutive cases of intrinsic brainstem lesions that were managed via an EEA, as well as the specific anatomy of each case. METHODS: All patients were treated in a single center by a multidisciplinary surgical team between 2015 and 2019. Before surgery, a dedicated anatomical analysis of the brainstem safe entry zone was performed, and proper surgical planning was carried out. Neurophysiological monitoring was used in all cases. Anatomical dissections were performed in three human cadaveric heads using 0° and 30° endoscopes, and specific 3D reconstructions were executed using Amira 3D software. RESULTS: All lesions were located at the level of the ventral brainstem. Specifically, one mesencephalic cavernoma, two pontine ca- vernomas, one pontine gliomas, and one medullary diffuse midline glioma were reported. Cerebrospinal fluid leak was the major complication that occurred in one case (medullary diffuse midline glioma). From an anatomical standpoint, three main safe entry zones were used, namely the anterior mesencephalic zone (AMZ), the peritrigeminal zone (PTZ, used in two cases), and the olivar zone (OZ). Reviewing the literature, 17 cases of various brainstem lesions treated using an EEA were found. CONCLUSIONS: To our knowledge, this was the first preliminary clinical series of intrinsic brainstem lesions treated via an EEA presented in the literature. The EEA can be considered a valid surgical alternative to traditional transcranial approaches to treat selected intra-axial brainstem lesions located at the level of the ventral brainstem. To achieve good results, surgery must involve comprehensive anatomical knowledge, meticulous preoperative surgical planning, and intraoperative neurophysiological moni- toring.

The effect of the volume of supra-inguinal injected solution on the spread of the injectate under the fascia iliaca: a preliminary study.

The fascia iliaca compartment is the compartment confined by the fascia iliaca (FI) and a muscular layer formed by the iliac- and psoas muscle. This compartment creates a virtual tunnel that contains the femoral nerve (FN), the obturator nerve (ON), and the lateral femoral cutaneous nerve (LFCN) of the lumbar plexus. In this pilot study, we aimed to determine the suggested volume needed to reach the three target nerves of the lumbar plexus (FN, ON, and LFCN) with a single-injection ultrasound-guided supra-inguinal fascia iliaca compartment (S-FICB). A computer tomography (CT scan)-guided step-up/step-down sequence was used to determine the suggested injection volume to target all three nerves. Subsequently, an anatomist blinded for the injected volume and CT findings, dissected the cadavers, and evaluated the spread of dye underneath the fascia iliaca. In total, seven pelvic areas of four cadavers were evaluated on CT scan and dissected. Distribution of dye underneath the FI in relation to the FN, ON, and the LFCN was recorded in all dissected cadavers. Combining CT and dissection findings, the suggested volume to reach the FN, ON, and LFCN with an S-FICB was 40 mL.

3D Models of Female Pelvis Structures Reconstructed and Represented in Combination with Anatomical and Radiological Sections.

We present a computer program designed to visualize and interact with three-dimensional models of the main anatomical structures of the female pelvis. They are reconstructed from serial sections of corpse, from the Visible Human project of the Medical Library of the United States and from serial sections of high-resolution magnetic resonance. It is possible to represent these three-dimensional structures in any spatial orientation, together with sectional images of corpse and magnetic resonance imaging, in the three planes of space (axial, coronal and sagittal) that facilitates the anatomical understanding and the identification of the set of visceral structures of this body region. Actually, there are few studies that analysze in detail the radiological anatomy of the female pelvis using three-dimensional models together with sectional images, making use of open applications for the representation of virtual scenes on low cost Windows® platforms. Our technological development allows the observation of the main female pelvis viscera in three dimensions with a very intuitive graphic interface. This computer application represents an important training tool for both medical students and specialists in gynecology and as a preliminary step in the planning of pelvic floor surgery.

The ventral route to intracranial aneurysm: from the origin towards modern transsphenoidal surgery. An historical review and current perspective.

OBJECTIVE: A review of the main studies that have explored the use of the ventral pathway for treatment of intracranial aneurysms,including the recent reported extended transsphenoidal approaches. METHODS: A comprehensive literature review was performed using the PubMed database. We recovered 48 cases of cerebral aneurysms, approached via the transcervical-transclival, transoral-transclival, transfacial-transclival ventral pathways and the extended transsphenoidal route. The overall rates of complications and surgical success were evaluated and compared for both traditional ventral and transsphenoidal approaches. RESULTS: For traditional routes, the overall complications and surgical success rates were 74% (26/35) and 87% (13115), respectively.For extended transsphenoidal approaches were 44% (4/9) and 78% (7 /9), respectively. CONCLUSION: Our paper is a reconnaissance of what has been done via "the anterior route" and a notification of the existence of this "surgical window': Present and future of cerebral aneurysm treatment is represented by the endovascular technique. A few selected cases in specialized centers, where transsphenoidal approaches with the aid of the endoscope are routinely performed,may be treated with such techniques alone or in combination with other different procedures. Further studies in large numbers of patients will be required to validate the full benefit of this approach.

Computed anatomical modelling of the optic pathway and oculomotor system using magnetic resonance imaging.

This study presents a computer-based tool for three-dimensional (3D) visualization of the optic pathway and oculomotor system using 3D high-resolution magnetic resonance imaging (MRI) datasets from a healthy subject. The 3D models were built as wireframe grids co-registered with MRI sections. First, 3D anatomical models were generated of the visual pathway from the eyeball to the primary visual cortex and of the cranial oculomotor nerves from the brain stem to the extrinsic eye muscles. Second, a graphical user interface allowed individual and group visualization, translation, rotation and zooming of the 3D models in different spatial positions simultaneously with MRI orthogonal cut planes. Educational and clinical applications are also discussed.

Dislodgment of continuous suprascapular nerve block catheters after physiotherapy: A cadaver study.

BACKGROUND: Continuous peripheral nerve blocks are commonly used for pain management. However, the incidence of catheter dislodgement or migration is unclear, and may be underestimated and underreported. Our objective was to assess suprascapular catheter tip positioning before and after routine simulated shoulder physiotherapy manipulation in an anatomical cadaver model. METHOD: Eight ultrasound-guided continuous suprascapular nerve block catheters were placed in cryopreserved fresh cadavers. Computed tomography (CT) confirmed the location of the catheter tip after injection of 1 ml of contrast medium. We performed a series of standardized shoulder movements during a simulated shoulder physiotherapy session in cadavers. Following this, we administered 1 ml of methylene blue through the catheters, and then performed anatomical dissections to accurately identify the location of the catheter tips and compare them to their placement prior to the 'physiotherapy'. RESULT: CT imaging confirmed the location of the catheter tips at the suprascapular notch in all cases. However, following physiotherapy, 2 catheters (25%) were found to have migrated - specifically, 1 was located in the supraspinatus muscle, and the other was located in the trapezius muscle. CONCLUSION: Our findings suggest that catheter dislodgement may occur in approximately 25% of cases following simulated physiotherapy manipulation. However, further research is needed to determine the read incidence of catheter dislodgement in clinical practice.

Serratus plane block: a novel ultrasound-guided thoracic wall nerve block.

We present a novel ultrasound-guided regional anaesthetic technique that may achieve complete paraesthesia of the hemithorax. This technique may be a viable alternative to current regional anaesthetic techniques such as thoracic paravertebral and central neuraxial blockade, which can be technically more challenging and have a higher potential side-effect profile. We performed the serratus block at two different levels in the midaxillary line on four female volunteers. We recorded the degree of paraesthesia obtained and performed fat-suppression magnetic resonance imaging and three-dimensional reconstructions of the spread of local anaesthetic in the serratus plane. All volunteers reported an effective block that provided long-lasting paraesthesia (750-840 min). There were no side-effects noted in this initial descriptive study. While these are preliminary findings, and must be confirmed in a clinical trial, they highlight the potential for the serratus plane block to provide analgesia following surgery on the thoracic wall. We suggest that this novel approach appears to be safe, effective, and easy to perform, and is associated with a low risk of side-effects.

Three-dimensional magnetic resonance image of structures enclosed in the spinal canal relevant to anesthetists and estimation of the lumbosacral CSF volume.

Three-dimensional (3D) image-reconstruction of structures inside the spinal canal certainly produces relevant data of interest in regional anesthesia. Nowadays, all hospital MRI equipment is designed mainly for clinical diagnostic purposes. In order to overcome the limitations we have produced more accurate images of structures contained inside the spinal canal using different software, validating our quantitative results with those obtained with standard hospital MRI equipment. Neuroanatomical 3D reconstruction using Amira software, including detailed manual edition was compared with semi-automatic 3D segmentation for CSF volume calculations by commonly available software linked to the MR equipment (MR hospital). Axial sections from seven patients were grouped in two aligned blocks (T1 Fast Field Eco 3D and T2 Balance Fast Field Eco 3D-resolution 0,65 x 0,65 x 0,65 mm, 130 mm length, 400 sections per case). T2 weighted was used for CSF volume estimations. The selected program allowed us to reconstruct 3D images of human vertebrae, dural sac, epidural fat, CSF and nerve roots. The CSF volume, including the amount contained inside nerve roots, was calculated. Different segmentation thresholds were used, but the CSF volume estimations showed high correlation between both teams (Pearson coefficient = 0.98, p = 0.003 for lower blocks; Pearson 0.89, p = 0.042 for upper blocks). The mean estimated value of CSF volume in lower blocks (L3-S1) was 15.8 + 2.9 ml (Amira software) and 13.1 +/- 1.9 ml (software linked to the MR equipment) and in upper blocks (T11-L2) was 21 +/- 4.47 ml and 18.9 +/- 3.5 ml, respectively. A high variability was detected among cases, without correlation with either weight, height or body mass index. Aspects concerning the partial volume effect are also discussed. Quick semi-automatic hospital 3D reconstructions give results close to detailed neuroanatomical 3D reconstruction and could be used in the future for individual quantification of lumbosacral CSF volumes and other structures for anesthetic purposes.

[Structure of the arachnoid layer of the human spinal meninges: a barrier that regulates dural sac permeability]. / Morfología de la lámina aracnoidea espinal humana. Barrera que limita la permeabilidad del saco dural.

OBJECTIVES: Drugs injected into the epidural space are known to penetrate the subarachnoid space by simple diffusion through the dural sac. We aimed to study the cellular ultrastructure of the arachnoid membrane and the type of intercellular junctions responsible for creating the barrier that regulates the passage of drugs through the dural sac in humans. MATERIAL AND METHODS: Fourteen tissue samples of arachnoid membrane were taken from 2 patients during procedures that required opening the lumbar dural sac. The samples were treated with glutaraldehyde, osmium tetroxide, ferrocyanide and acetone, and then embedded in resin. Ultrathin sections were stained with lead citrate for examination by transmission electron microscopy. RESULTS: The arachnoid membrane was 35 to 40 microm thick. The outer surface contained neurothelial cells (dural border cells) along the subdural compartment, while the internal portion was made up of a plane 5 to 8 microm thick with 4 to 5 arachnoid cells overlapping to form a barrier layer. The intercellular spaces on this plane were 0.02 to 0.03 microm wide; the arachnoid cells were bridged by specialized junctions (desmosomes and other tight junctions). CONCLUSIONS: Structural features of the arachnoid cells provide a barrier within the human dural sac. They occupy only the internal portion of the arachnoid membrane. Specialized intercellular junctions explain the selective permeability of this membrane.

Anatomical description of anaesthetic spread after deep erector spinae block at L-4. / Estudio anatómico de la distribución del volumen administrado tras bloqueo en el plano profundo del erector espinal a nivel lumbar.

INTRODUCTION: Thoracic erector spinae plane (ESP) block is now used for postoperative analgesia. However, although reports of lumbar ESP have been published, the anesthetic spread and mechanism of action of this technique remains unclear. We describe the lumbar ESP block technique and evaluate the spread of 20ml of solution administered at the level of the transverse process of L4 in a cadaver model. METHODS: Observational study after 12 lumbar ESP blocks at L4 on a fresh cadaver model (6 bilaterally). The spread of 20ml of injected contrast solution was assessed by computed tomography in all 6 samples. Four of the samples were evaluated by anatomical study, 2 by plane dissection, and 2 others were frozen and cut into 2-2.5cm axial slices. RESULTS: The injected solution spread from L2 to L5 in a cranio-caudal direction in the erector spinae muscle, reaching the facet joints medially and the thoracolumbar fascia laterally. In 33% of cases the solution did not spread anterior to the transverse process; in 51%, spread was minimal and did not affect the corresponding spinal nerves, and in 2 samples (16%), spread was extensive and reached the corresponding spinal nerves. CONCLUSIONS: Lumbar ESP at L4 always acts on the posterior branches of the spinal nerves, but seldom spreads to the paravertebral space to block the spinal nerve.

Representations of hindlimb digits in rat dorsal root ganglia.

The distribution in dorsal root ganglia of neurones that innervate the distal tips of the hindlimb digits in the rat were mapped after subcutaneous injections of the fluorescent tracers Fast Blue, Diamidino Yellow, and Fluoro-Gold into different digits. Three-dimensional reconstruction was used to describe the intraganglionic distribution of neurones labelled from different digits. Labelled neurones were found mainly in the L3-L5 ganglia. The distribution in ganglia and the number of neurones labelled from each digit varied considerably between cases, but mean numbers of labelled neurones were similar for the different digits. Neurones in L3 tended to innervate medial digits and neurones in L5 tended to innervate lateral digits, but most neurones from any digit were found in L4. Although overlap was considerable, the three-dimensional reconstruction showed tendencies of neurones to be distributed in restricted territories within the dorsal root ganglia. This was especially clear in ganglion L5, where digit IV was found to be represented more rostrally than digit V. The results indicate that primary afferent neurones that innervate the hindlimb digits are represented by a crude rostrocaudal somatotopic organisation both among and within lumbar dorsal root ganglia.

Fast blue and diamidino yellow as retrograde tracers in peripheral nerves: efficacy of combined nerve injection and capsule application to transected nerves in the adult rat.

Capsule application of Diamidino Yellow (DY) to the cut end of the sciatic nerve immediately followed by capsule application of Fast Blue (FB) resulted in approximately 95% double-labelled dorsal root ganglion neurones (DRGn) and motoneurones (Mn). Nerve injection of DY followed either immediately or 2 months later by capsule application of FB resulted in approximately 90% double-labelled DRGn and Mn, indicating that DY and FB label similar populations of DRGn and Mn, and that insignificant DY fading occurred during this period. Inversing the order of application, however, i.e. nerve injection of FB followed immediately by capsule application of DY, resulted in double labelling in only approximately 10% of the DRGn and Mn. These percentages increased to 70% of the DRGn and 60% of the Mn when the FB injection was followed 1 or 2 months after by the DY application, indicating that DY uptake is blocked by recent administration of FB. The results indicate that DY and FB might be useful for sequential labelling before and after nerve injury as a tool to investigate the accuracy of sensory and motor regeneration.

Efficacy of the fluorescent dyes Fast Blue, Fluoro-Gold, and Diamidino Yellow for retrograde tracing to dorsal root ganglia after subcutaneous injection.

The present study was designed to investigate the efficacy of the fluorescent dyes Fast Blue (FB), Fluoro-Gold (FG), and Diamidino Yellow (DY) for retrograde tracing of lumbar dorsal root ganglia after their subcutaneous injection into different hindlimb digits. Injections of equal volumes (0.5 microl) of 51% FB or 2% FG resulted in similar mean numbers of sensory neurones labelled by each tracer. Injection of equal volumes (0.5 microl) of FB or FG in a single digit followed 10 days later by a second injection of the same volume of 5% DY into the same digit resulted in similar mean numbers of labelled sensory neurones for each of the three tracers. Furthermore, on average, 75% of all the FB-labelled cells and 74% of all FG-labelled cells also contained DY. Repeating the same experiment with an increased volume of DY (1.5 microl) resulted in an increase in the mean number of double-labelled profiles to 82 and 84% for FB and FG, respectively. The results show that FB, FG and DY label similar numbers of cutaneous afferents and that a high level of double labelling may be obtained after sequential injections in digits. These properties make them suitable candidates in investigations where a combination of tracers with similar labelling efficacies is needed.

Sciatic and femoral nerve sensory neurones occupy different regions of the L4 dorsal root ganglion in the adult rat.

The topographical distribution of sciatic and femoral nerve sensory neuronal somata in the L4 dorsal root ganglion of the adult rat was mapped after retrograde tracing with one or two of the dyes Fast Blue, Fluoro-Gold, or Diamidino Yellow. The tracers were applied to the proximal transected end of either nerve alone, or from both nerves in the same animal using separate tracers. Three-dimensional reconstructions of the distribution of labelled neurones were made from serial sections of the L4 dorsal root ganglion which is the only ganglion that these two nerves share. The results showed that with little overlap, femoral nerve neurones distribute dorsally and rostrally whereas sciatic nerve neurones distribute medially and ventrally. This finding indicates the existence of a somatotopical organisation for the representation of different peripheral nerves in dorsal root ganglia of adult animals.

Desplazamiento de los catéteres para bloqueo continuo del nervio supraescapular tras fisioterapia: estudio en cadáveres / Dislodgment of continuous suprascapular nerve block catheters after physiotherapy: A cadaver study

Antecedentes: Los bloqueos continuos de los nervios periféricos son de uso común para el manejo del dolor. Sin embargo, la incidencia de desplazamiento o migración del catéter es dudosa, pudiendo ser infraestimada y no denunciarse. Nuestro objetivo fue evaluar el posicionamiento del extremo del catéter supraescapular antes y después de la manipulación rutinaria y simulada mediante fisioterapia de hombro en un modelo cadavérico anatómico.MétodoSe colocaron 8 catéteres para bloqueo continuo y ecoguiado del nervio supraescapular en cadáveres frescos criopreservados. La tomografía computarizada (TC) confirmó la localización de la punta del catéter tras la inyección de 1ml de medio de contraste. Realizamos una serie de movimientos estandarizados de hombro durante una sesión simulada de fisioterapia de hombro en cadáveres. Tras ello, administramos 1ml de azul de metileno a través de los catéteres, y seguidamente realizamos disecciones para identificar con precisión la localización de las puntas del catéter y compararlas con su localización previa a la «fisioterapia».ResultadosLa imagen de la TC confirmó la localización en todos los casos de los extremos del catéter en la muesca supraescapular. Sin embargo, tras la fisioterapia, se encontró que 2 catéteres (25%) habían migrado y, en especial, uno fue localizado en el músculo supraespinoso y el otro en el músculo trapecio.ConclusiónNuestros hallazgos sugieren que el desplazamiento del catéter puede producirse en cerca del 25% de los casos tras la manipulación mediante fisioterapia simulada. Sin embargo, es necesaria más investigación para determinar la incidencia real de los desplazamientos del catéter en la práctica clínica. (AU)

Background: Continuous peripheral nerve blocks are commonly used for pain management. However, the incidence of catheter dislodgement or migration is unclear, and may be underestimated and underreported. Our objective was to assess suprascapular catheter tip positioning before and after routine simulated shoulder physiotherapy manipulation in an anatomical cadaver model.MethodEight ultrasound-guided continuous suprascapular nerve block catheters were placed in cryopreserved fresh cadavers. Computed tomography (CT) confirmed the location of the catheter tip after injection of 1ml of contrast medium. We performed a series of standardized shoulder movements during a simulated shoulder physiotherapy session in cadavers. Following this, we administered 1ml of methylene blue through the catheters, and then performed anatomical dissections to accurately identify the location of the catheter tips and compare them to their placement prior to the ‘physiotherapy’ResultCT imaging confirmed the location of the catheter tips at the suprascapular notch in all cases. However, following physiotherapy, 2 catheters (25%) were found to have migrated - specifically, 1 was located in the supraspinatus muscle, and the other was located in the trapezius muscle.ConclusionOur findings suggest that catheter dislodgement may occur in approximately 25% of cases following simulated physiotherapy manipulation. However, further research is needed to determine the read incidence of catheter dislodgement in clinical practice. (AU)

Cerebrospinal fluid volume and nerve root vulnerability during lumbar puncture or spinal anaesthesia at different vertebral levels.

Cerebrospinal fluid (CSF) and nerve root volumes within the lumbosacral dural sac were estimated at various vertebral levels, in an attempt to determine any possible relevance to the incidence of nerve root trauma during lumbar puncture or spinal anaesthesia. Magnetic resonance images from seven patients were studied. Volumes were calculated by semi-automatic threshold segmentation combined with manual editing of each slice. The mean dural sac volume from S1 to T12 was 42.8±5.8 ml and the mean CSF volume 34.3±5.1 ml with the mean root volume being 10.4±2.2 cm(3). The mean CSF volume per vertebral segment ranged from 4.3±0.7 ml at L5, to 5.8±2.5 ml at L1, with high inter-individual variability. The mean root volume ranged from 0.6±0.1 cm(3) at L5 to 2.4±0.5 cm(3) at T12. The conus medullaris was located at L1 in four of the five patients scanned at upper lumbar levels, and at the lower border of L2 in the other. Vulnerability to nerve root damage was expressed as the Vulnerability Index (%), being defined as the ratio of root volume to dural sac volume (CSF volume + root volume). The value ranged between 7 and 14% at L5, increasing rostrally to 30 to 43% at T12. Caution is obviously required in high punctures to avoid contact with the conus medullaris, but the cauda equina is also vulnerable to contact with more caudal punctures and had a Vulnerability Index of about 25% at L4, that increased rostrally.

An anatomical study of the intradural space.

We have previously postulated that it is possible to accidentally insert an epidural needle or catheter into the substance of the dura during attempted epidural block, creating an intradural space. It appears that injection of local anaesthetic into an intradural space leads to an initially inadequate neuraxial block but further doses may produce an extensive life-threatening block. In the laboratory, 54 samples of human thoraco-lumbar dura were obtained from six cadavers and prepared for scanning electron microscopy. Images from these dehydrated specimens were assessed for the presence of spaces within the dura, and attempts were made to insert an epidural catheter, under optical microscopy, into the substance of the dura in 32 cases. Electron microscopy revealed the concentric laminae that compose the dura and the presence of artefactual spaces between some of these. It was possible to insert an epidural catheter into the substance of the dura in eight specimens, creating intradural spaces which remained following catheter removal. If this represents the clinical situation, it may help to explain previously reported cases of atypical neuraxial block and their associated radiological findings.