Unlike the cervical spine where several vertebrae have distinctly identifiable anatomy (C1, C2, C7) and the number of vertebrae are nearly constant, the vertebrae of the lumbar spine are all reasonably similar in shape/configuration but variable in number. It is fairly common to have four or six lumbar type vertebral bodies instead of five and transitional lumbosacral vertebral bodies are also common. Thus, establishing a clearly communicated enumeration is the first key step in assessment of radiologic anatomy and levels of pathology.

The next key concept is that anatomic alignment on both coronal and sagittal images is critically important, but can vary substantially between recumbent and weight-bearing positions. Since most cross-sectional anatomy imaging exams are performed in a recumbent position,* comparison to standing plain radiographs is a key, but often overlooked, step. Changes in alignment on standing can lead to exacerbated spinal canal, neural foraminal, and/or lateral recess stenosis with resultant neural compression. Clues to motion on recumbent images such as gas within a disc space at a level after attempted fusion or facet joint effusions, but can also be seen at levels that do not show changes in alignment with position/weight-bearing.

Other global anatomic features to assess are the width of the spinal canal, which can be congenitally narrow in the setting of short pedicles, the extent of epidural fat, and global changes of the marrow such as proliferation or replacement or osteopenia.

Two of the three most commonly implicated specific culprits of axial low back pain, the intervertebral disc and facet joint, are located in the lumbar spine (the third is the SI joint-but that topic is complicated as discussed next). It is important to understand the normal anatomy and how to differentiate expected age-related changes from pathology that can be associated with pain. This is a point of confusion and will be addressed in the separate courses and questions.

*there is some difference also between prone and supine imaging, but a larger difference between recumbent (prone or supine) and standing.

References

Fardon DF, Lumbar disc nomenclature: version 2.0: recommendations of the combined task forces of the North American Spine Society, the American Society of Spine Radiology, and the American Society of Neuroradiology. Spine J 2014;14(11):2525-45

Bones, thecal sac, cauda equina, paraspinal tissues.

Show axial with central, subarticular, neural foraminal, and far lateral areas.

Radicular pain, or manifestations of spinal canals stenosis...Much imaging centers around spinal canal, lateral recess, and neural foraminal narrowing.

Numerous proposed pain generators of the low back are depicted, although not by any means exhaustively. Essentially every structure of the lower back with nociceptive innervation has been proposed to be a possible pain generator at some point by someone. However, each structure is not a proven common/realistic pain generator and the mechanism(s) of pain generation from any individual structure is often not well-understood.

The lumbar intervertebral disc contains a central nucleus pulposis and peripheral annulus fibrosis. The annulus is contiguous with the fibrocartilaginous vertebral endplates, the ALL, and the PLL. In a healthy disc, the nucleus pulposis is well-hydrated. Upon application of an axial load, it acts in hydraulic fashion to resist deformation and transfer pressure circumferentially to the annulus fibrosis. Desiccation of the nucleus pulposis results in decreased hydraulic response to axial load and the annulus bears a greater burden of this load. There is decreased T2 signal intensity of the nucleus pulposis and blurring between it and the annulus fibrosis.

Some degree of nucleus pulposis dehydration is seen with normal aging. However, with internal disc disruption (likely secondary to an endplate fracture), a series of biochemical events can ensue, resulting in accelerated dessication, annular fissures, and ultimately marked dessication, height loss, and/or disc protrusion/extrusion. These changes can be accompanied by Modic endplate changes. There is extensive literature assessing the specific imaging findings (decreased T2 signal, height loss, annular fissures, bulges, herniations, Modic changes) to discogenic pain as determined by provocation discography, although this is a complicated topic and beyond the scope of this section. The innervation of the intervertebral disc and endplates as well as the clinical implications are briefly discussed below.

The three main components of the intervertebral disc are water, proteoglycans, and collagen. The nucleus pulposis consists of up to 90% water. A proteoglycan backbone, shown as a spiral in this case contributes to viscosity/cohesion and has glycosaminoglycan side-chains (green) with negatively charges that attract water. Overall, the nucleus pulposis is a hydrated, but viscous, substance that helps withstand axial load. With age, it becomes less hydrates, acquires more collagen, and looks increasing like the annulus fibrosis (less T2 hyperintense).

The annulus fibrosis has a more prominent collagenous component. Collagen fibrils are organized into fibers that concentrically encircle the nucleus pulposis.

The nucleus pulposis is encapsulated by the annulus fibrosis and cartilaginous endplate. The annulus has two main components: an outer component that bridges the ring apophyses and is essentially functions as the 'intervertebral ligament.' The inner portion contains cartilage fibers that are contiguous with those of the cartilaginous endplate. The endplate has cartilaginous and osseous components. Given the association with the annulus fibrosis, the cartilaginous endplate may be best considered a portion of the intervertebral disc rather than vertebral body.

One mechanistic theory of IDD is that the initiating event is an endplate fracture. Such a fracture could occur with forces experienced in daily life (repeated heavy lifting, falls, etc.) without major trauma and may not be symptomatic. This fracture initiates degradation of the nucleus pulposis. Possible mechanisms include violation of the normal immune-privileged status of the nucleus pulposis through exposure to the vertebral marrow or other loss of homeostasis within the nucleus pulposis.

The water content and viscosity decrease. Thus, the hydrostatic resistance to axial load decreases with stress shifted to the annulus fibrosis. This leads to height loss and disc bulge. Once a fissure extends into the annulus fibrosis, there is IDD. A fissure may reach nerve endings, inciting pain, and may act as a conduit for a disc herniation.

The illustration shows degradation of the nucleus pulposis. With decreased water content and viscosity of the nucleus pulposis, there is decreased ability to resist axial load. A greater burden of axial load resistance is shifted to the annulus fibrosis. This results in disc height loss, decreased integrity of the annulus fibrosis collagen fibers, and potentially a disc bulge.

Demonstration of a grade III annular fissure (Dallas Classification) that reaches the outer 1/3 of the annulus fibrosis. There is also an extrusion of degraded nuclear material. Non-degraded nuclear material would resist extrusion as it retains greater viscosity.

The basivertebral nerve courses alongside the basivertebral vein in the posterior midline vertebral body. After entering the vertebral body, it ramifies and innervates the vertebral endplates. This is the rationale for the more recent interventional practice of basivertebral nerve ablation for treatment of pain arising from Modic changes.

The sinuvertebral nerve and autonomics do not provide discrete 1-to-1 innervation to spinal levels, but instead form plexuses around the vertebral bodies.

The lumbar facet joints are true synovial joints with a joint capsule, synovium, articular cartilage, and subchondral bone. The orientation can vary level-by-level, person-by-person, and even side-to-side, The joint space is often curved in the axial plain and appears over a range of projection angles on fluoroscopy. Facet joints contribute to stability and can be a source of axial back pain. Joint hypertrophy can be associated with ligamentum flavum hypertrophy and can contribute to neural foraminal, lateral recess, and spinal canal stenosis. Effusions within the facet joint can be associated with degenerative change and sometimes segmental instability. This fluid can also lead to synovial cysts, which may exacerbate stenoses.

The L1-L2 through L4-L5 facet joints receive dual innervation from the medial branch at the respective level and the level above. The medial branch typically takes a course that includes the groove at the junction of the superior articular process and transverse process, constituting a target for diagnostic medial branch blocks and RF ablation for facet joint pain. Nerve endings branch, including innervation of the facet joint capsule. Some treatment modalities have attempted to target (ablate) nerve endings over the capsule as an alternative to the medial branches.

The neural foramen are a keyhole shaped opening that contain the exiting ventral ramus/dorsal root ganglia surrounded by epidural fat. Key determinants of neural foraminal size or stenosis include the intervertebral disc, facet joint, and spinal alignment. The height of the neural foramen is partially maintained by the height of the intervertebral disc and disc height loss contributes to neural foraminal narrowing in the craniocaudal dimension. A disc bulge or protrusion can efface

Radicular (or radiculomedullary) arteries vary in location, but are typically located within the superior half of the foramen. Inadvertent access and injection of such an artery if contributing to the anterior spinal artery can result in catastrophic spinal cord infarct and paralysis.

These are best assessed by a combination of sagittal and axial images.

Example of infraneural injection, stenoses of different etiologies, conjoined nerve root in inferior portion of the foramen....

The ligaments of the lumbar spine are complex, frequently misunderstood structures.

First, the outer rim of the annulus fibrosis bridge the ring apophyses of adjacent vertebral bodies, effectively functioning as ligaments of an 'interverbral joint.' This 'ligament' opposes multiple types of vertebral body motion; only with axial loading does it function in concert with the nucleus pulposis. Again, the fibers are closely associated with the ALL and PLL.

The ALL is layered, including a deep layer with fibers bridging two adjacent vertebral bodies across the disc level and longer superficial fibers. It is difficult to differentiate the ALL from the tendons of the crura of the diaphragm. This ligament opposes hyperextension.

The PLL is also layered and is relatively deficient laterally at the disc-space level. This accounts for the predilection of off-midline/paracentral located disc herniations.

Ligaments of the posterior elements include the ligamentum flavum, interspinous ligament, and supraspinous ligament. The facet joint capsule is not truly a ligament, but does contribute to stability.

The ligamentum flavum is an elastic structure bridging adjacent lamina. It separates the epidural space ventrally from the retrodural space (see below) dorsally. It also forms the ventral facet joint capsule.

The ventral and middle portions of the interspinous ligament are truly an interspinous ligament, spanning adjacent spinous processes. The dorsal interspinous ligament extends from the superior surface of the spinous processes and courses dorsally to join the supraspinous ligament. The ventral portion splits into paired structures usually separated by fat, but sometimes by an interspinous bursa.

The supraspinous ligament usually terminates caudally at the L5 level without crossing the L5-S1 level. In the lumbar spine, it is largely comprised of paraspinal tendons and aponeurotic tissue and the designation as a bon a fide ligament could be debated.

Finally, the iliolumbar ligament is a complex structure, but the primary component extends from the transverse process of the lowest lumbar-type vertebral body (usually, but not always, L5) to the ilium. Some radiologists use this to help designate a transitional vertebral body as lumbar type.

Reference.

Bogduk N. Clinical and radiological anatomy of the lumbar spine 5th edition.

Carinno JA et al. Effect of spinal segment variants on numbering vertebral levels at lumbar MR imaging. Radiology 2011;259(1):196-202

Depending on configuration of the transitional segment, there may be increased stress of the level above the segment due to the morphology, location, and potential partial fusion of the transitional segment with the sacrum. For instance, this may be associated with pain from a facet joint above the segment or a pseudoarticulation of the transitional segment transverse process with the saral ala. The somewhat controversial term 'Bertolloti Syndrome' may be used to refer to axial low back pain related to a transitional lumbsacral segment, but does not denote one specific pain generator.

Beyond the basic anatomy discussed above, there is lots more clinically relevant detail about the anatomy and clinical significance that can facilitate evaluation of cross-sectional imaging. Much of this centers around the differentiation of typical age-related change from pain-generating pathology, difficultly correlating specific pain generators with pain due to overlapping and variable pain patterns, and variable association of certain imaging findings with pain. There are also misconceptions and common knowledge gaps in the evaluation of post-operative status and trauma. For these reasons and more, interpretation of spine examinations can be deceptively difficult.

One advanced concept that is useful to know is the existence of the retrodural space of Okada. This space is located dorsal to the ligamentum flavum and has potential to communicate with multiple adjacent spaces including bilateral facet joints, adjacent pars interarticularis defects, interspinous bursae, and the dorsal epidural space. The retrodural space allows spread of fluid or inflammatory change, may be inadvertently injected after early loss-of-resistance with interlaminar injections, may be opacified during pain interventional procedures targeting the facet joints/pars defects, and is occasionally identified on MRI.