The epidural space is defined as the area within the vertebral canal between the dura mater of the spinal cord and the lining of the vertebral canal.1 The epidural space contains fat and the internal ventral vertebral venous plexus. The veterinary literature contains little information regarding the epidural space and its contents, especially the fibrous attachments of the dura mater to the vertebral bodies and their clinical relevance. For example, a fibrous band connects the filum terminale to the dorsal aspect of the vertebral bodies.2 Fibrous connections also surround the L6 and L7 spinal nerve roots and intervertebral foramina in large-breed dogs.3 Those structures may protect dogs from motion-induced spinal nerve injury, or conversely, they may have pathological consequences following degenerative changes.3
Humans have numerous MVLs including the anterior, posterior, and lateral fibrous attachments that traverse the epidural space and connect the dura mater to the posterior surface of vertebral bodies.4–11 In the anterior epidural space, the fibrous attachments of the anterior surface of the dura mater and the posterior aspect of the vertebral body are known as Hofmann ligaments.10–12 Those ligaments have been implicated in the topography of pathological processes involving the vertebrae13 and the distribution of extruded IVD material within the epidural space.9,14
The primary purpose of the study reported here was to determine whether there is an anatomic connection between the external ventral surface of the dura mater of the spinal cord and the dorsal surface of the vertebral bodies, which forms the ventral boundary of the epidural space, in dogs. An additional goal was to describe gross and histologic characteristics of naturally occurring vertebral neoplasms, which were topographically affected by the MVL, in 2 dogs. Finally, diagnostic imaging findings for 2 dogs with HNPE were examined and used as a basis for proposing the MVL as an anatomic explanation for the bilobed shape of the extruded IVD material commonly observed in such dogs.
Materials and Methods
The study was purely descriptive. Vertebral column or surgically extirpated epidural tissue specimens were grossly and histologically or cytologically evaluated for 3 groups of dogs: 6 neurologically normal dogs, 2 dogs with vertebral neoplasms, and 2 dogs with HNPE. For the neurologically normal dogs, the presence and strength of the MVL were visually and qualitatively assessed. All dogs were privately owned and euthanized for routine necropsy for reasons unrelated to the study. Dogs were euthanized by IV injection of an overdose of a pentobarbital solution following the acquisition of owner consent. To avoid bias, the only selection criterion for dogs was that they be neurologically normal.
The vertebral column (C1 to sacrum) was harvested, and the paravertebral musculature was removed. Then, a dorsal laminectomy was performed by use of a Ruskin rongeur or autopsy sawa along the entirety of the vertebral column as described.15 The vertebral columns from 3 dogs were used for gross evaluation of fresh tissues, and the vertebral columns from the other 3 dogs were fixed in neutral-buffered 10% formalin for histologic examination.
For the 3 fresh specimens, the meninges at the cervicomedullary junction were grasped with Brown-Adson forceps and lateral traction was applied with enough force to expose the spinal nerves, which enabled transection of the spinal nerve roots or nerves. Subsequently, dorsal traction of the meninges was used to elevate the spinal cord from the epidural space. Subjective assessment of the force needed to pull the spinal cord out of the epidural space was noted as the attachment of the ventral dura mater to the dorsal aspect of the vertebral bodies was observed. This process was repeated for each subsequent pair of spinal nerves or roots and spinal cord segment.
For the formalin-fixed specimens, a variable-speed rotary tool and a metal cutting wheelb was used to make 2-cm-thick transverse sections through the spinal cord and vertebral body at the cranial, middle, and caudal portions of each of the following vertebrae: C2 through C7, T5, T10, and L4. Sections were routinely decalcified until the bone was easily sectioned with a double-edged grossing blade. Sections were processed for histologic evaluation by the use of 2 methods. For 1 method, sections were made with care taken to not alter the anatomic relationship between the spinal cord and meninges within the epidural space. For the other method, sections were made while traction was applied to the dorsal aspect of the meninges to artificially enlarge the ventral epidural space. Then, liquefied agar was poured into the enlarged ventral epidural space and allowed to gelatinize to maintain the altered anatomic position of the meninges. Following gross sectioning, all specimens were routinely processed, embedded in paraffin, sliced into 5-μm-thick sections, and stained with H&E stain for histologic evaluation.
For the 2 dogs with vertebral neoplasms that extended into the epidural space, tissue specimens of the neoplasms and adjacent vertebral column were processed and histologically evaluated in the same manner as the tissue specimens obtained from the neurologically normal dogs. Both dogs were examined because of paraparesis. One dog underwent MRI of the vertebral column by use of a 1.5-T MRI unit.c T1-weighted and T2-weighted sequences were obtained in the sagittal and transverse planes before and after administration of gadopentetate dimeglumined (0.1 mmol/kg [0.05 mmol/lb], IV). Computed tomographye of the vertebral column was performed for the other dog. Both dogs had diagnostic imaging results that were suggestive of a neoplasm, and the owners elected euthanasia (IV injection of a pentobarbital overdose) and necropsy.
For the 2 dogs with HNPE, material extirpated from the epidural space during decompressive ventral slot surgery was cytologically evaluated. Diagnostic images were also reviewed. Both dogs had acute onset of rapidly progressive nonambulatory tetraparesis and clinical signs that were consistent with myelopathy between C1 and C5. The dogs did not exhibit signs of pain when the neck was manipulated. For both dogs, the cervical portion of the vertebral column from the caudal aspect of the skull to T3 was imaged with a 3-T MRI unit.f Briefly, each dog was anesthetized and positioned in dorsal recumbency. T1-weighted and T2-weighted sequences were obtained as described for one of the dogs with a vertebral neoplasm. Each dog also underwent 3-D TOF MRV by use of the following parameters: repetition time, 22 milliseconds; time to echo, 4.4 milliseconds; averages, 1; slice thickness, 0.8 mm; flip angle, 18°; field of view, 160 × 160 mm; and imaging matrix, 320 × 218. To overcome in-plane saturation, data were obtained as a multiple thin-slab acquisition in which 2 or 3 slabs were prescribed on the basis of the length of the vertebral column that was imaged. The 3-D TOF MRV data were viewed as MIP reconstructions in a plane that was parallel to the craniocaudal course of the internal ventral vertebral venous plexus. Diagnosis of HNPE was made on the basis of the presence of previously described imaging characteristics,16–19 and each dog underwent a ventral slot surgical procedure to extirpate the material that was extruding into the epidural space and compressing the spinal cord.
All tissue specimens and diagnostic images evaluated in the study were acquired as part of a routine diagnostic workup for clinical cases with the owners' consent. Consequently, the study was exempt from institutional animal care and use committee review.
Results
Neurologically normal dogs
The 6 neurologically normal dogs included 3 mixed-breed dogs, 1 Border Collie, 1 German Shepherd Dog, and 1 pit bull-type dog that ranged in age from 2 to 6 years old. There were 4 neutered males and 2 spayed females.
In all 6 dogs, there was a thin, fibrous attachment within the vertebral canal between the ventral aspect of the dura mater and the dorsal aspect of the vertebral bodies, which was visible when sufficient traction was applied to the dorsal aspect of the dura mater to elevate the spinal cord out of the vertebral canal. The fibrous tissue attached to the median bony ridge of the dorsal aspect of the vertebral bodies near the dorsal longitudinal ligament. This attachment appeared as a continuous thin band (width, < 0.1 to 0.2 mm) of fibrous tissue when the spinal cord was elevated from the vertebral canal and was present from C3 to L7 or the sacrum. It was most clearly visible and robust from C3 through C5 or C6. In fact, subjectively more force was required to disrupt the attachment at that location than elsewhere along the vertebral column. Moreover, the attachment was stronger along the body of each vertebrae than over the IVDs, where it was easily disrupted. In the cervical portion of the vertebral canal, the fibrous attachment provided almost as much resistance as did the spinal nerve roots and often required sharp dissection during removal of the spinal cord from the epidural space. In the thoracic and lumbar portions of the vertebral canal, the fibrous attachment was not as visible as it was in the cervical portion, and gentle traction often resulted in the easy removal of multiple segments of the spinal cord from the epidural space.
Gross examination of formalin-fixed transverse sections of the vertebral column and spinal cord revealed that the fibrous attachment was nearly indistinguishable as a separate structure from the ventral aspect of the dura mater. It appeared as a slight thickening of the ventral external surface of the dura mater adjacent to the dorsal longitudinal ligament and dorsal midline of the vertebral bodies. When slight traction was applied to the dorsal aspect of the dura mater, the fibrous attachment became visible as a meshwork of fibrous stands or as 1 or 2 thin fibrous strands that spanned from the ventral external surface of the dura mater to the dorsal longitudinal ligament and attached to the dorsal midline of the vertebral bodies (Figure 1)
Photographs of transverse sections of the vertebral column obtained just cranial to the midbody of the C3 (A), C4 (B), and C5 (C) vertebrae of a neurologically normal 3-year-old pit bull-type dog. In all 3 sections, 25-gauge hypodermic needles were used to apply sufficient traction on the dorsal aspect of the dura mater to elevate the spinal cord from the epidural space so that the MVL (arrow) could be visually observed. Notice that the MVL consists of fibrous bands, which extend from the ventral external surface of the dura mater to the dorsal longitudinal ligament and bony ridge at the midline of the dorsal surface of the vertebral body. Bar = 0.5 cm.
Citation: Journal of the American Veterinary Medical Association 255, 6; 10.2460/javma.255.6.687
When the formalin-fixed specimens were histologically evaluated, the fibrous connection between the dura mater and dorsal midline of the vertebral bodies was not visible in transverse sections in which the spinal cord was dorsally distracted within the epidural space and maintained in that position by gelatinized agar. However, thin fibrous strands coursing in a dorsoventral direction within the ventral epidural space were evident. In transverse sections in which the anatomic relationship of the dura mater and spinal cord within the epidural space was not altered, a fibrous connection between the ventral external surface of the dura mater and dorsal surface of the vertebral bodies was readily evident. Specifically, the fibrous strands blended into the collagen of the dorsal longitudinal ligament located on the surface of the dorsal midline of the vertebral bodies (Figure 2)
Photomicrographs of transverse sections of the ventral aspect of the spinal cord, ventral epidural space, ventral internal vertebral venous plexus (asterisk), and dorsal aspect of the vertebral body obtained just cranial to the midbody of the C3 (A), C4 (B), and C5 (C) vertebrae of a neurologically normal 2-year-old Border Collie. The MVL (arrow) spans from the ventral external surface of the dura mater (black arrowhead) to the dorsal longitudinal ligament (open arrowhead) and bony ridge at the midline of the dorsal surface of the vertebral body. H&E stain; bar = 500 μm. Insets represent lower-magnification photomicrographs of the section in each panel. H&E stain; bar = 500 μm.
Citation: Journal of the American Veterinary Medical Association 255, 6; 10.2460/javma.255.6.687
Dogs with vertebral neoplasms
The 2 dogs with vertebral neoplasms included a 10-year-old 25-kg (55-lb) spayed female Bassett Hound and a 6-year-old 32-kg (70-lb) castrated male Labrador Retriever. The neoplasms of both dogs had a similar gross appearance, although one was characterized as an osteosarcoma and the other as a plasma cell neoplasm. Both neoplasms expanded from a vertebral body into the ventral epidural space. On transverse sections, the neoplasms had a bilobed shape, which variably occupied the ventral epidural space and caused compression of the internal vertebral venous plexus and spinal cord. In both dogs, the cleft between the lobes of the neoplasm was centered over the dorsal midline of the affected vertebral body or bodies (Figure 3) Fibrous tissue was grossly observed to span from the ventral aspect of the dura mater to the ventral midline of the vertebral canal. Histologically, the thin fibrous attachment to the ventral aspect of the dura mater was not evident, which suggested that it was disrupted during tissue processing and sectioning; however, 1 to 2 thin fibrous strands attached to the dorsal longitudinal ligament and the midline bony ridge of the dorsal aspect of the vertebral body remained intact (Figure 4)
Photograph of a transverse section of the vertebral column obtained at the cranial aspect of the body of C5 in a 6-year-old Labrador Retriever with a plasma cell neoplasm (red dotted line) that extended from C5 into the ventral epidural space. Notice that the portion of the neoplasm within the ventral epidural space has a bilobed shape, which was attributed to the presence of the MVL (arrow) that extended from the ventral external surface of the dura mater to the dorsal longitudinal ligament (arrowhead). Bar = 0.5 cm. Inset—Postcontrast transverse T1-weighted MRI image with fat saturation obtained at the same level as the tissue specimen in the photograph. Notice the strong homogeneous contrast enhancement of the neoplastic tissue, which accentuates the bilobed shape of the neoplasm within the ventral epidural space and facilitates visualization of the MVL and dorsal longitudinal ligament.
Citation: Journal of the American Veterinary Medical Association 255, 6; 10.2460/javma.255.6.687
Photomicrographs (A and C) and photograph (B) of a transverse section of the vertebral column obtained at the level of T7 in a 10-year-old Basset Hound with a vertebral osteosarcoma (asterisks) that extended from T7 into the ventral epidural space. A—The portion of the neoplasm extending into the ventral epidural space had a bilobed appearance, and the lobes were delineated by the ventral aspect of the dura mater (open arrowhead), thin fibers of the MVL (arrow), and dorsal longitudinal ligament (black arrowhead). H&E stain; bar = 500 μm. B—The bilobed shape of the neoplasm, MVL, and dorsal longitudinal ligament could also be grossly appreciated. Bar = 500 μm. C—Lower-magnification photomicrograph of the tissue section in panel A, in which the section is similar in size to the gross section in panel B. H&E stain; bar = 500 μm. Notice that the MVL appears to create an anatomic barrier between the ventral external aspect of the dura mater and dorsal longitudinal ligament that restricts the neoplastic tissue from migrating across the dorsal midline of the vertebral body.
Citation: Journal of the American Veterinary Medical Association 255, 6; 10.2460/javma.255.6.687
Dogs with HNPE
The 2 dogs with HNPE included an 8-year-old 9.7-kg (21.3-lb) mixed-breed dog and a 7-year-old 21-kg (46.2-lb) Springer Spaniel. The MRI findings were similar for both dogs. Briefly, T2 hyperintense material centered over an IVD expanded into the ventral epidural space, which caused focal compression of the spinal cord. On transverse T2-weighted images, the ventral aspect of the extradural material had a bilobed appearance, and the cleft between the lobes was centered on the midline at the area where the fibrous attachment between the ventral aspect of the dura mater and dorsal aspect of the vertebral bodies and IVDs was observed in the neurologically normal dogs (Figure 5) The nucleus pulposus that remained within the annulus fibrosis at the site of spinal cord compression was decreased in size but had an MRI signal intensity that was similar to that of adjacent unaffected IVDs. For both dogs, MIP reconstruction of 3-D TOF MRV sequences revealed a signal intensity consistent with flowing blood in the area of the ventral internal vertebral venous plexus. The diameter of the ventral internal vertebral venous plexus in the affected area was similar to that in the unaffected regions. In sequences obtained after contrast administration, contrast medium within the ventral internal vertebral venous plexus outlined focal extradural material in the ventral epidural space that was isointense to hypointense relative to the spinal cord.
Representative MRI images of the vertebral column of 2 dogs with HNPE. A—Transverse T2-weighted MRI image acquired at the C3–4 IVD space of an 8-year-old mixed-breed dog. The herniated nucleus pulposus (arrow) is hyperintense and has a bilobed shape with the location of the cleft between the lobes consistent with the location of the MVL. Inset—Photomicrograph of a cytologic preparation of a portion of the material extirpated from the ventral epidural space during a decompressive ventral slot surgical procedure. Notice that the cytologic characteristics are most consistent with notochordal cells. Wright-Giemsa stain; bar = 40 μm. B—Transverse T2-weighted MRI image acquired at the C4–5 IVD space of a 7-year-old Springer Spaniel. Similar to the dog in panel A, the herniated nucleus pulposus (arrow) is hyperintense and has a bilobed shape owing to the presence of the MVL. Inset—Dorsal plane MIP reconstruction of 3-D TOF MRV sequences for the dog of panel B, which depicts normal blood flow in the ventral internal venous plexus at the site of the herniated nucleus pulposus at the C4–5 IVD space (ie, between the arrowheads).
Citation: Journal of the American Veterinary Medical Association 255, 6; 10.2460/javma.255.6.687
Both dogs underwent spinal cord decompression by means of a standard ventral slot procedure. Following resection of the dorsal annulus and disruption of the dorsal longitudinal ligament, a gray viscous material was easily identified, removed, and submitted for cytologic analysis. Of particular note was the fact that the extruded nucleus pulposus did not become visible until the dorsal ligament was resected, and hemorrhage from the ventral internal venous plexus was not encountered. Therefore, the extirpated material must have been located within the ventral epidural space. The extirpated material was cytologically characterized as notochordal cells and chondrocytes with a background of RBCs, which was consistent with nucleus pulposus.
Discussion
Epidural fat and the internal vertebral venous plexus exist within the epidural space. Additionally, the spinal nerve roots exit the dura mater to course through the epidural space toward their respective intervertebral foramina. The epidural space also contains the MVL (which is referred to as the Hofmann ligament in humans) and fibrous lining of the vertebral canal.7,20,21 That lining has been termed the peridural membrane or periosteum22; however, it may lack a layer of osteogenic cells and might not meet the definition of a true periosteum. The peridural membrane can be considered analogous to a periosteal lining.21
Prior to the present study, the veterinary literature lacked a description of the MVL. In fact, only the peridural membrane had been summarily described, with focus mainly on its relationship with the L6 and L7 spinal nerves as they exit their respective intervertebral foramina.3 Results of the present study expanded knowledge regarding the anatomy of the epidural space by establishing the presence of an MVL that spans from the ventral external surface of the dura mater to the dorsal aspect of the vertebral bodies and IVDs in neurologically normal dogs as well as dogs with vertebral neoplasms and dogs with HNPE.
In humans, MVLs are thin, discontinuous attachments between the dura mater and posterior longitudinal ligament in the anterior epidural space. The most robust ligaments are located on the midline, whereas finer MVLs are located more laterally.11,21,23 Although MVLs are predominately observed in the lumbar portion of the vertebral column, they are also present in the cervical portion and poorly represented in the thoracic portion of the vertebral column.23 For the dogs evaluated in the present study, a single MVL was observed in all portions of the vertebral column. It formed a continuous attachment between the ventral aspect of the dura mater and dorsal aspect of the vertebral bodies and IVDs along the midline and was most robust in the cervical portion of the vertebral column, particularly between C3 and C5 or C6. Unfortunately, similar to the MVLs of humans, the small size and inherent fragility of the MVL of dogs made it difficult to observe the structure in specimens following routine processing for histologic examination.10,20 Nevertheless, the conspicuous gross appearance of the MVL in fresh specimens validated its existence and likely provided a better assessment of the ligament's physical characteristics than would assessment of its histologic properties.
Descriptions of anatomic structures are intrinsically valuable, but it is generally the function of those structures that is of clinical importance. Despite debate, it is likely that the MVL functions to help maintain a close anatomic relationship between the vertebrae and spinal cord during development and growth.11 The MVL also anchors the meninges of the spinal nerves to the vertebrae and IVDs, which may prevent stretching of those nerves, and thereby pain, during movement of the vertebral column.21 Conversely, the anchoring of the spinal nerve meninges to the vertebral column by the MVL can also contribute to signs of pain by not allowing the spinal nerve roots to move in the event of IVD herniation.24 The functional significance of the MVL in dogs was beyond the scope of the present study; rather, we simply focused on describing the gross and histologic appearance of the MVL in neurologically normal dogs and dogs with naturally occurring vertebral neoplasms as well as on MRI sequences in dogs with HNPE.
The exact anatomic location of herniated material in dogs with HNPE has been debated in the literature. For the 2 dogs with HNPE evaluated in the present study, MIP reconstruction of 3-D TOF MRV sequences and gross observation during surgery confirmed that the herniated material was present extradurally in the ventral epidural space. Similarly, for the dogs with vertebral neoplasms examined in the present study, the gross and histologic appearance of the affected portions of the vertebral column revealed that the neoplasms extended into the ventral epidural space. Interestingly, although the pathogenesis of vertebral neoplasms and HNPE is vastly different, both types of lesions had a bilobed appearance, which suggested that the lesion boundaries were dictated by something within the epidural space rather than any intrinsic characteristics of the disease process. Moreover, epidural lesions in dogs with sterile inflammation of fat, empyema, and hemorrhage also have a bilobed shape.25–27 The vertebral canal and spinal cord represent anatomic constraints for lateral and dorsal expansion of pathological lesions within the epidural space, but the bilobed appearance of such lesions with the interlobular cleft centered over the midline would seem to be best explained by the presence of an MVL.
In human patients, many types of vertebral neoplasms have a bilobed appearance on MRI sequences, and that topography is attributed to the MVL and peridural membrane rather than neoplastic-specific characteristics.13 In fact, the conspicuity of the midline MVL increases as the anterior epidural space becomes increasingly occupied by neoplastic tissue.13 A similar phenomenon was observed in the present study in both the 2 dogs with vertebral neoplasms and 6 neurologically normal dogs when tension was applied to the dorsal aspect of the dura mater.
Results of the present study indicated that an MVL exists in dogs. It creates an anatomic barrier, which causes pathological lesions within the ventral epidural space to adopt a bilobed shape regardless of the pathogenic process. The anatomic barrier created by the MVL may also account for the lateralization of extruded degenerative IVD material by preventing its migration across the midline.14 In dogs, lateralization of herniated degenerative IVD material often occurs in the cervical portion of the vertebral column.28 It is likely that the MVL prevents movement of the extruded degenerative IVD material across the midline when it migrates cranial or caudal to the IVD space. Prevention of extruded IVD material from crossing the midline causes it to migrate into the intervertebral foramen, which may explain why dogs with degenerative IVD herniation in the cervical portion of the vertebral column often develop compression of cervical nerve roots (ie, root signature signs). Dogs with degenerative IVD herniation can also develop epidural hemorrhage, which is often unilateral likely owing to the presence of the MVL.29 In humans, the MVL can attach to the internal vertebral venous plexus, and inadvertent trauma to the MVL during surgical procedures can lead to hemorrhage from the internal vertebral venous plexus.30 In an analogous manner, the occurrence of epidural hemorrhage in dogs with IVD herniation may be the result of disruption of the MVL attachment, which was immediately adjacent to the medial aspect of the ventral internal vertebral plexus in the cadaveric specimens of the present study. The MVL may also have direct effects on the spinal cord. In dogs, the dura mater is under tension and stretched longitudinally along the length of the vertebral column.31 It is likely that the MVL anchors the dura mater to the vertebral column. The denticulate ligaments that attach the spinal cord to the arachnoid layer and dura mater apply less stretching force on the spinal cord than does the tension applied to the dura mater.31 It is possible that disruption of the MVL, as occurs with traumatic vertebral fractures or luxations, may result in dura mater tension being transferred to the spinal cord, which may exacerbate spinal cord injury. Congenital vertebral malformation-induced kyphosis or scoliosis may likewise lead to the transfer of tension and force from the dura mater to the spinal cord.
Findings of the present study suggested that the anatomic constraints within the vertebral canal of dogs, including the MVL, cause the predictable bilobed topography of lesions in the ventral epidural space independent of the pathogenic mechanism. Further anatomic studies of dogs are warranted to detail the lining of the epidural space, MVL and its potential variations along the length of the vertebral canal, spinal nerves and roots, and the spinal cord so that the distribution and disease processes within the vertebral canal can be elucidated.
ABBREVIATIONS
| HNPE | Hydrated nucleus pulposus extrusion |
| IVD | Intervertebral disk |
| MIP | Maximum-intensity projection |
| MRV | Magnetic resonance venography |
| MVL | Meningovertebral ligament |
| TOF | Time-of-flight |
Footnotes
810 autopsy saw, Stryker Corp, Kalamazoo, Mich.
4000 high-performance rotary tool, Dremel, Mount Prospect, Ill.
1.5-Tesla, Brivo MR 355 Inspire, General Electric Medical Healthcare, Milwaukee, Wis.
Magnevist, Bayer HealthCare LLC, Whippany, NJ.
Lightspeed 16, General Electric Medical Healthcare, Milwaukee, Wis.
3.0-Tesla MRI, Siemens Skyra, Erlangen, Germany.
References
1. Fletcher TF. Spinal cord and meninges In: Evans HE, de Lahunta A, eds. Miller's anatomy of the dog. 4th ed. St Louis: Elsevier Saunders, 2012;589–610.
2. Fletcher TF, Kitchell RL. Anatomical studies on the spinal cord segments of the dog. Am J Vet Res 1966;27:1759–1767.
3. Breit S, Giebels F & Kneissl S. Foraminal and paraspinal extraforaminal attachments of the sixth and seventh lumbar spinal nerves in large breed dogs. Vet J 2013;197:631–638.
4. Barbaix E, Girardin MD, Hoppner JP, et al. Anterior sacrodural attachments—Trolard's ligaments revisited. Man Ther 1996;1:88–91.
5. Connor MJ, Nawaz S, Prasad V, et al. The posterior epidural ligaments: a cadaveric and histological investigation in the lumbar region. ISRN Anat 2013;2013:424058.
6. Geers C, Lecouvet FE, Behets C, et al. Polygonal deformation of the dural sac in lumbar epidural lipomatosis: anatomic explanation by the presence of meningovertebral ligaments. AJNR Am J Neuroradiol 2003;24:1276–1282.
7. Loughenbury PR, Wadhwani S, Soames RW. The posterior longitudinal ligament and peridural (epidural) membrane. Clin Anat 2006;19:487–492.
8. Plaisant O, Sarrazin JL, Cosnard G, et al. The lumbar anterior epidural cavity: the posterior longitudinal ligament, the anterior ligaments of the dura mater and the anterior internal vertebral venous plexus. Acta Anat (Basel) 1996;155:274–281.
9. Scapinelli R. The meningovertebral ligaments as a barrier to the side-to-side migration of extruded lumbar disc herniations. Acta Orthop Belg 1992;58:436–441.
10. Wadhwani S, Loughenbury P, Soames R. The anterior dural (Hofmann) ligaments. Spine (Phila Pa 1976) 2004;29:623–627.
11. Wiltse LL, Fonseca AS, Amster J, et al. Relationship of the dura, Hofmann's ligaments, Batson's plexus, and a fibrovascular membrane lying on the posterior surface of the vertebral bodies and attaching to the deep layer of the posterior longitudinal ligament. An anatomical, radiologic, and clinical study. Spine (Phila Pa 1976) 1993;18:1030–1043.
12. Tardieu GG, Fisahn C, Loukas M, et al. The epidural ligaments (of Hofmann): a comprehensive review of the literature. Cureus 2016;8:e779.
13. Schellinger D. Patterns of anterior spinal canal involvement by neoplasms and infections. AJNR Am J Neuroradiol 1996;17:953–959.
14. Schellinger D, Manz HJ, Vidic B, et al. Disk fragment migration. Radiology 1990;175:831–836.
15. Vandevelde M, Higgins R, Oevermann A. General neuropathology. In: Veterinary neuropathology: essentials of theory and practice. Ames, Iowa: Wiley, 2012;1–37.
16. Manunta ML, Evangelisti MA, Bergknut N, et al. Hydrated nucleus pulposus herniation in seven dogs. Vet J 2015;203:342–344.
17. Dolera M, Malfassi L, Marcarini S, et al. Hydrated nucleus pulposus extrusion in dogs: correlation of magnetic resonance imaging and microsurgical findings. Acta Vet Scand 2015;57:58.
18. Beltran E, Dennis R, Doyle V, et al. Clinical and magnetic resonance imaging features of canine compressive cervical myelopathy with suspected hydrated nucleus pulposus extrusion. J Small Anim Pract 2012;53:101–107.
19. De Decker S, Fenn J. Acute herniation of nondegenerate nucleus pulposus: acute noncompressive nucleus pulposus extrusion and compressive hydrated nucleus pulposus extrusion. Vet Clin North Am Small Anim Pract 2018;48:95–109.
20. Parkin IG, Harrison GR. The topographical anatomy of the lumbar epidural space. J Anat 1985;141:211–217.
21. Wiltse LL. Anatomy of the extradural compartments of the lumbar spinal canal. Peridural membrane and circumneural sheath. Radiol Clin North Am 2000;38:1177–1206.
22. Ansari S, Heavner JE, McConnell DJ, et al. The peridural membrane of the spinal canal: a critical review. Pain Pract 2012;12:315–325.
23. Scapinelli R. Anatomical and radiologic studies on the lumbosacral meningo-vertebral ligaments of humans. J Spinal Disord 1990;3:6–15.
24. Spencer DL, Irwin GS, Miller JS. Anatomy and significance of fixation of the lumbosacral nerve roots in sciatica. Spine (Phila Pa 1976) 1983;8:672–679.
25. Theobald A, Dennis R, Beltran E. Imaging diagnosis—spontaneous subperiosteal vertebral hemorrhage in a Greyhound. Vet Radiol Ultrasound 2014;55:420–423.
26. Carrera I, Sullivan M, McConnell F, et al. Magnetic resonance imaging features of discospondylitis in dogs. Vet Radiol Ultrasound 2011;52:125–131.
27. Murata D, Miura N, Iwanaga T, et al. CT and MRI imaging diagnosis of epidural idiopathic sterile pyogranulomatous inflammation in a dog spinal canal. J Vet Med Sci 2012;74:913–915.
28. Ryan TM, Platt SR, Llabres-Diaz FJ, et al. Detection of spinal cord compression in dogs with cervical intervertebral disc disease by magnetic resonance imaging. Vet Rec 2008;163:11–15.
29. Mateo I, Lorenzo V, Foradada L, et al. Clinical, pathologic, and magnetic resonance imaging characteristics of canine disc extrusion accompanied by epidural hemorrhage or inflammation. Vet Radiol Ultrasound 2011;52: 17–24.
30. Chen R, Shi B, Zheng X, et al. Anatomic study and clinical significance of the dorsal meningovertebral ligaments of the thoracic dura mater. Spine (Phila Pa 1976) 2015;40:692–698.
31. Tunituri AR. Elasticity of the spinal cord dura in the dog. J Neurosurg 1977;47:391–396.
