3.19 Cross-sectional imaging in spinal disorders

Disc Degeneration

Disc degeneration is a multifactorial disease. There is a general misconception to consider degeneration as a part of the normal ageing process but many factors including mechanical, genetic, nutritional, and traumatic causes have all been implicated in its aetiology. Similar changes have been described in asymptomatic young individuals, further evidence that genetic factors play a significant role (Figure 3.19.1).

A wide spectrum of imaging techniques has been used to image low back pain. Plain film changes of degenerative disc disease (DDD) are a late finding. Computed tomography (CT) changes can be appreciated earlier, but only when the disc has bulged beyond its normal margins or when intradiscal gas or osteophytes have developed. Due to its ability to detect changes in the water content of the disc, magnetic resonance imaging (MRI) plays a principal role in identifying the degenerate disc and in differentiating other causes of pain; however, there are no reliable findings that help to differentiate the painful from the painless degenerate disc. Recent advances in MRI like functional T₂ mapping can also evaluate disc function.

The adolescent disc demonstrates three different anatomical areas on T₂-weighted (T2W) water-sensitive MRIs. There is a black outer zone of annulus fibrosus, a bright inner zone of the nucleus pulposus, and an intermediate zone, histologically representing a transition between the regular lamellae of the annulus and the more loosely packed nucleus. With increasing age, the ability of the nucleus to imbibe water decreases and there is a gradual loss of its bright signal. The early changes of disc degeneration parallel this change, but occur at a faster rate. Lumbar disc degeneration assessment is graded on the basis of MRI signal intensity on T2W midsagittal fast spin echo images, and distinction between nucleus and annulus, disc structure, and height. A grade 1 disc is homogenous, hyperintense, has normal height with a clear distinction between nucleus and annulus. A grade 5 disc is inhomogeneous, hypointense, with a collapsed disc space with loss of distinction between nucleus and annulus.

Cadaveric studies have shown that annular disruption is the decisive factor in disc degeneration. Annular tears, particularly radial tears are not necessarily a consequence of ageing. Annular tears can be radial, concentric, or transverse. A specific type, with a hyperintense round lesion surrounded by a black annulus has been called a high signal zone (HSZ) (Figure 3.19.2). There is some evidence that discs with HSZs are more likely to be symptomatic at discography. Other authors however, disagree. Annular tears are more conspicuous following gadolinium enhancement. On serial MRI of annular tears in lumbar discs, it has been suggested that hyperintensity on T2W MRIs and enhancement of annular tears do not indicate that the tear is acute in nature. With progressive DDD there is loss of vertical height, disc material bulges in all directions, and the facet joints sublux. The combination of these changes can narrow the spinal canal, lateral recesses, or exit foramina and cause nerve root compression.

The adjacent vertebral bodies also show changes in form of signal alteration (Table 3.19.1).

Type 3 endplate changes can be seen on plain films. Type 1 changes are non-specific and simply indicate marrow oedema (Figure 3.19.3).When present on both sides of a disc, which demonstrates low signal on T2W images, increased marrow water can usually be attributed to disc degeneration rather than tumour or infection. Discitis is also associated with peridiscal marrow oedema however, in contradistinction to DDD, the infected disc has high signal on T2W images (Figure 3.19.4). In some cases of severe disc degeneration, the nucleus can become liquefied, and also return high signal on T₂. In these cases, differentiation from discitis can be more difficult. As many of the infections are low grade and inflammatory indices are often normal or only minimally elevated, enhancement of the vertebral body endplates following administration of gadolinium-labelled diamino-tetraethyl-pentaacetic acid (Gd-DTPA) may assist, but it is non-specific and can also occur after uncomplicated surgery. Simultaneous enhancement within the endplate, posterior annulus, and vertebral body itself is more specific for infection. The pattern of endplate changes may provide a clue, in that fatty type 2 changes are unlikely to persist adjacent to an actively infected disc. In other cases, biopsy may be required to exclude infection (Figure 3.19.5).

Annular disc degeneration may lead to annular tear and disc herniation. Disc herniation is most common at the L4/5 and L5/S1 levels. Plain radiographs have no role in the investigation of a patient with sciatica. Increasing use of non-ferromagnetic metalwork has improved the diagnostic capability of MRI after surgery as well. In the postfusion patient with radiculopathy, where implanted metal may interfere with both CT and MR interpretation, CT myelography may still have a role.

On MRI, disc prolapse is classified as disc herniation or disc bulge. If disc displacement in axial plane beyond the edges of the ring apophysis is less than 50% it is disc herniation and if more it is termed as disc bulge. Disc bulge is regarded as a degenerative process distinct from disc herniation.

Disc herniation is further subdivided on the basis of circumference, shape, and location. On the circumferential basis, localized displacement in axial plane beyond the edges of the ring apophysis is of two types: focal, if it is less than 25% and broad based, when the displacement is 25–50%.

On the basis of the shape it is classified as: protrusion, if the greatest distance in any plane, between the edges of the disc material beyond the disc space is less than the distance between the edges of the base of the herniated disc in the same plane (Figure 3.19.6); otherwise it is termed as extrusion. Extrusion can be referred to as sequestration if the displaced disc material has no continuity with the parent disc. Migration is a term used when the disc material is displaced away from the site of extrusion, irrespective of sequestration (Figure 3.19.7).

On the basis location, in the axial plane, it is classified as: central, subarticular (lateral recess region), foraminal (pedicle), and extra foraminal (far lateral). In sagittal plane it is categorized as discal, pedicular, supra- and infrapedicular. Nerve sheath tumours and facet synovial cysts can occasionally mimic disc extrusion or sequestration on MRI and may pose a diagnostic challenge.

A number of studies have emphasized that disc prolapse as revealed by MRI have been noted in asymptomatic individuals. In addition, the grade of disc degeneration or prolapse does not predict the clinical outcome. Furthermore, histologically proven degenerate discs with substantially decreased amount of nuclear material can still produce normal images on MRI.

The patients with new or persistent symptoms following surgery (failed back syndrome) pose a difficult diagnostic dilemma. If symptoms persist or recur then images should be inspected to confirm that surgery was undertaken at the correct level and side. Thus, identification of lumbosacral transitional vertebra (LSTV) on MRI is very important. Disc bulge, protrusion, and spinal canal stenosis is nine times more common at the interspace immediately above LSTV. L5 can be readily identified on axial MRI as iliolumbar ligament always arises from the transverse processes of L5. If these points have been confirmed, persistent or new low back pain or recurrent sciatica following discectomy can be attributed to recurrent or residual disc material, epidural fibrosis or haematoma, discitis, arachnoiditis, or abscess (Figures 3.19.8 and 3.19.9).

Plain films are still used to assess spinal fusion, especially to assess metal discontinuity or fracture and displacement. Flexion and extension views can be used to assess vertebral stability. Cross-sectional CT imaging has largely superseded plain radiography as it can provide more detailed information. Implanted metal is not a contraindication to MRI or CT per se; however, depending on the metal used, paramagnetic artefacts, distortion, and beam hardening can obscure anatomic detail. In comparison to the steel implants, titanium metal implants cause fewer artefacts on CT as they have a low x-ray attenuation coefficient and also have less ferromagnetic affect and cause less susceptibility artefact on MRI. Ultrasound is utilized to diagnose, delineate the size and location of postoperative collections, and aspirate superficial collections. Where there is clinical suspicion of postoperative discitis, contrast-enhanced MRI is the imaging investigation of choice; however, in the early stages it may be difficult to discriminate between septic and aseptic discitis. Imaging findings must be correlated with the clinical picture and biochemical markers.

In persistent or recurrent radicular symptoms, the differential diagnosis is between recurrent disc material, neural irritation from epidural granulation tissue, and, rarely, residual disc material. It has been proposed that contrast-enhanced MRI is superior to MRI, CT, and myelography in discriminating recurrent disc from scar tissue. Where enhancement is used, it has been suggested that ionic MR contrast media, in comparison to non-ionic media, diffuses less readily into disc material and provides greater contrast with scar. The addition of fat-suppressed T₁ images and obtaining images immediately after the contrast injection may improve scar visualization. Residual disc may only show rim enhancement due to surrounding granulation tissue. The degree of enhancement of scar tissue is also dependent on its age and reaches a maximum by the first year. Disc enhancement may also be seen if imaging is delayed.

Gadolinium should not be regarded as mandatory to diagnose recurrent disc material. In equivocal cases, gadolinium enhancement does help, although it is important to appreciate that vascularized discs may enhance and dense fibrosis may not. The presence of epidural scar tissue on MRI does not necessarily indicate a poor outcome. Disc degeneration is a risk factor for recurrent herniation irrespective of its volume.

The MRI appearance of epidural haematoma depends on its age. In the early phase, the signal characteristic is that of fluid—low on T₁ and high on T₂. As it begins to deteriorate, methaemoglobin release causes the signal on T₁ to increase. Finally, haemosiderin deposition causes the signal on T₂ to become low. In most cases, haematoma comprises several of these stages.

Other potential MRI findings in the failed back syndrome include arachnoiditis, seromata, and, rarely, pseudomeningocele. Arachnoiditis is diagnosed by the identification of one of several patterns. The most specific is clumping and agglutination of multiple nerve roots. Another pattern, termed the ‘empty sac’, occurs when the nerve roots adhere to the thecal sac itself, leaving an apparently empty cerebrospinal fluid space. This pattern can be seen in asymptomatic individuals, emphasizing the importance of clinical correlation.

The most common tumour encountered is the vertebral haemangioma. As these lesions contain both fat and fluid elements, its MRI appearances are usually typical; the lesion is bright on both T₁ and T₂ images (Figure 3.19.10). Metastases and myeloma may have variegated marrow appearance on MRI (Figure 3.19.11) (Table 3.19.2).

Spondylolisthesis is classified as degenerative, isthmus (spondylolytic) and dysplastic, traumatic, and pathological. Degenerative spondylolisthesis is commonly noted at L4/5 and isthmic is more often at L5/S1 (Figure 3.19.13). Axial loaded MRI may show dynamic degenerative spondylolisthesis, which may be missed on plain radiography or conventional MRI. A facet joint effusion of 1.5mm or more at L4/5 is suggestive of degenerative spondylolisthesis even if anterior translation is not noted on supine lumbar spine MRI. On MRI, indirect features of pars spondylosis are widening of the canal size and posterior vertebral wedging and are helpful in diagnosing pars defect in absence of spondylolisthesis. A full discussion of infection, tumour, ankylosing spondylitis, and spondylolisthesis is beyond the scope of this chapter.

Vertebral body biopsy is a safe procedure when carried out with appropriate care. Coagulopathy should be excluded prior to the procedure. CT or fluoroscopic guided biopsy can be performed. Fluoroscopic procedures are quicker than CT guided procedures, but CT provides more accurate needle placement for smaller lesions (Figure 3.19.14).

CT-guided core biopsy was found to be 93.3% accurate with highest accuracy rates for malignant lesions and false negative for benign, inflammatory, pseudotumoural and systemic pathologies. A negative biopsy in neoplastic lesions must be further confirmed with an open biopsy. Reported complications including infection, haematoma, and self-limiting neurapraxia are rare.

Percutaneous access to the lumbar and thoracic discs is readily achieved via posterolateral and cervical discs via an anterolateral route under fluoroscopic control.

Statistics on lost man-hours and health costs in the patients with chronic disabling back pain show no evidence of abating despite a proliferation of surgical procedures and fusion devices. Part of the difficulty is patient selection, as routine imaging cannot confirm an abnormality as the definitive source of pain. This is where discography has a role. Discography can provide both anatomical and more importantly ‘physiological’ information on the source of pain, though false negatives and positives are the subject of much controversy (Figure 3.19.15).

Discography is most useful at differentiating painful from non-painful discs and should be regarded as a preoperative examination. Discography is valuable in selected patients with moderate loss of nuclear signal intensity and no other MRI features of DDD. Lesser roles include differentiating scar tissue from recurrent disc and confirming disc containment prior to percutaneous surgery. Pressure-controlled manometric discography can distinguish asymptomatic discs with grade 3 (Dallas Discogram Scale) annular tears from symptomatic discs.

It is of two types, intradiscal electrothermal therapy (IDET) and percutaneous intradiscal radio frequency thermocoagulation (PIRFT). In the former, heat is generated electrically and in the latter through radiofrequency. These modalities are used in treatment of painful annular tears by coagulating the inflammatory tissue and the nerve endings, thus ablating pain. The treatment, however, is not permanent and, if successful, has to be repeated. In IDET a catheter is passed into the nucleus pulposus such that it lies along the inner margin of the annulus whereas in PRIFT the catheter is placed into the centre of the disc. Initial non-randomized studies suggested a favourable outcome for IDET but not for PRIFT. Various randomized controlled trials (RCTs) have suggested both favourable and unequivocal outcomes. A recent study, however, has shown that with application of strict selection criteria (patients with mild disc degeneration, annular tear proven on imaging, and pain on low-pressure discography), superior outcomes can be achieved with IDET.

Prolapsed discs that have failed to respond to conservative measures can be treated by a variety of intra- and extradiscal measures. Intradiscal treatment includes chemonucleolysis, percutaneous nucleotomy, and laser nucleotomy. All involve instrumentation of the nucleus pulposus and its removal by chemical dissolution, suction, or heat.

Chemonucleolysis helps in reducing turgidity in the nucleus and decreasing pressure on the nerve root. Significant improvement of outcome and reduction in recurrence rate has been demonstrated with an addition of 1000U of an intradiscal chymopapain following transforaminal endoscopic discectomy. Percutaneous laser disc decompression has made vaporization of a small area of the nucleus pulposus possible.

An alternative means of percutaneous disc excision is automated percutaneous lumbar discectomy. Dynamic stabilization system is useful to prevent progression of initial lumbar degenerative disc disease after nucleotomy for symptomatic prolapsed disc. According to NICE guidance (2006), the potential candidates for disc decompression using coblation are patients with contained disc leading to back and leg pain. Percutaneous endoscopic transforaminal lumbar discectomy (PELD) are also being performed for the herniated discs and forminoplastic PLED is considered to be an efficacious treatment option for soft migrated disc herniation.

Facet joint arthritis is easily seen on axial CT or MRI sections. The appearances mimic changes of osteoarthritis elsewhere, with joint-space irregularity, osteophytes, and subarticular sclerosis. It is frequently identified in the asymptomatic population. There are no firm clinical signs confirming the facet as the source of the patient’s symptoms. Injection of local anaesthetic and steroid into the facet has been used as both a diagnostic test and a therapeutic measure (Figure 3.19.16). There is moderate evidence of long- and short-term relief of lumbar facet joint pain by use of an intra-articular facet joint injection, medial nerve blockade, and neurotomy or rhizolyis. Occasionally, small synovial cysts can be identified arising from the facet joint. These can be intra- or extraspinal. Intraspinal lesions may present with neural compression (Figure 3.19.17). Imaging-guided cyst aspiration can be performed under fluoroscopic, CT, and MRI guidance. An interlaminar approach under fluoroscopic guidance has been used to puncture an intraspinal zygoapophyseal cyst to alleviate the radicular symptoms.

Each facet receives dual innervations from the median branches above and below. Median branch blockade is an alternative to facet blockade provided that both branches are injected. The target point is the elbow between the transverse and the superior articular process. Functional status improvement was noted in 82% of the patients with local anaesthetic nerve root block of the lumbar facet joint irrespective of steroid use. Rhizolysis (radiofrequency neurotomy) is similar to medial branch blockade but achieves more long-lasting results. Like IDET, this treatment is not permanent. If a satisfactory response to intra-articular facet injection is achieved, more long-term pain relief is gained by medial nerve neurotomy by radiofrequency ablation or cryoneurolysis.

Multilevel disc degeneration and multiple root symptoms are common in older age groups. Images obtained in the recumbent position may underestimate the presence or degree of root impingement. Selective root blockade provides a means whereby an individual root can be anaesthetized and the subsequent effect on symptoms determined (Figure 3.19.18).

Percutaneous vertebroplasty involves the injection of cement into a collapsed vertebral body, mainly for pain relief. Patient selection for both procedures should be performed with a multidisciplinary approach including spinal surgeon and radiologist. The principal indication is painful osteoporotic collapse. Indications also include malignant collapse and expansive intravertebral haemangioma. Vertebroplasty can be considered in patients with intractable pain due to vertebral fractures. It is contraindicated in coagulopathy. VERTOS and INVEST are two current multicentre RCTs that have yet to confirm a more favourable outcome with percutaneous vertebroplasty than conservative treatment.

The technique involves the injection of polymethylmethacrylate (PMMA) into collapsed vertebral bodies. The vertebral body is cannulated either by a direct posterolateral approach or via the transpedicular route. Once needles are positioned, cement is injected under pressure to support the vertebral body. Complications include extravasation of cement, which may lead to nerve or thecal compression (Figure 3.19.19).

A variation on this procedure is balloon kyphoplasty. This procedure is used to correct an angular deformity resulting from vertebral collapse. Following vertebral cannulation, a Teflon balloon is dilated under pressure within the vertebral body. Once the deformity has been corrected, the balloon is deflated and removed and the cavity is filled with cement.

Out of the two percutaneous vertebral augmentation techniques, vertebroplasty offers comparable pain relief but is safer as often only a unilateral approach is utilized, with fewer requirements of PMMA and without risk of adjacent level fracture.