Physical Address

304 North Cardinal St.
Dorchester Center, MA 02124

Infections of the Spinal Column


Spinal infection frequently injures the vertebral bodies, intervertebral discs, paraspinal soft tissues, epidural space, meninges, and spinal cord. Clinically, it is notoriously difficult to differentiate spinal infection from degenerative processes, noninfective inflammatory disorders, and spinal neoplasms. Spinal infections are a common cause of morbidity and mortality, especially in immunocompromised patients.

Epidural and intradural extramedullary infections are most often caused by pyrogens. Intramedullary infections are usually caused by viral agents. Infections in each spinal compartment present distinct imaging features, permitting differential diagnosis on the basis of characteristic imaging patterns.

PYOGENIC DISCITIS-OSTEOMYELITIS

Pyogenic discitis-osteomyelitis is infection of the vertebral bones and the adjacent discs. It is also referred to as discitis and vertebral osteomyelitis or as pyogenic spondylitis.

Epidemiology

Vertebral osteomyelitis presently accounts for 2% to 4% of all pyogenic bone infection. Its incidence appears to be increasing, possibly because of the increasing age of the population, increased intravenous drug abuse, AIDS, reactivation of latent infections, overuse of antibiotics, drug-resistant pathogens, new pathogens, worsening of socioeconomic conditions, and biologic war/terrorism. Other predisposing factors include malnutrition, immune compromise, chronic steroid use, diabetes mellitus, malignancy, chronic alcoholism, renal failure, implantation of intravascular devices, and recent spinal surgery.

Males are affected more frequently than females in a ratio of 1.5 to 3:1. Discitis is more common in children especially before age 4 years. After childhood, spinal infection shows a bimodal age distribution with a smaller peak in the second decade and a dominant peak in the fifth decade.

Clinical Presentation

The neurologic deficits depend on the site and extent of infection, the type and virulence of the causative agent, and the host's ability to contain or resist the infection. The onset of symptoms may be indolent, with only back pain, malaise, and discomfort. There may be a history of recent infection elsewhere in the body. The patient may complain of back pain persisting over months to years, malaise, fever, anorexia, muscular spasm, stiffness, rigidity, weakness, fatigue, dysphagia, weight loss, and night sweats. Typically, motion aggravates the pain of spinal infection and rest ameliorates it. Children may show irritability and refuse to walk, sit, or stand.

Physical examination often discloses local tenderness, subcutaneous flank mass, and spinal deformity, such as increased thoracic kyphosis. Neurologic examination reveals signs of nerve root compression, including radiculopathy, signs of meningeal irritation, and neurologic deficits, such as lower extremity weakness, loss of reflexes, and paraplegia. Collapse of the vertebral body and/or epidural inflammatory tissue may compress neural structures and their blood supply, leading to impaired cord function or infarction. Cervical spondylitis can extend anteriorly and progress to retropharyngeal abscess and mediastinitis, with esophageal dysfunction. Thoracic spondylitis can cause mediastinitis, empyema, and pericarditis. Lumbar spondylitis can cause peritonitis and subdiaphragmatic abscess. Cardiac and respiratory signs and symptoms vary according to the severity of the involvement.

Untreated infection may result in permanent spinal deformity with wedging of the vertebral bodies, angular kyphosis (gibbus), and scoliosis. Severe infection may cause permanent neurologic deficits, particularly paraplegia and tetraplegia. Extension of the infection to the psoas musculature may lead to transient or permanent hip contractures.

Chronic spinal infection may present a more confusing clinical picture. The pain is less severe. The onset of paresis of the extremities is gradual, with signs and symptoms of cord and/or cauda compression but no fever. Generally, pain is less severe.

Laboratory findings also vary with the grade of infection and the specific infective agent but typically include elevated erythrocyte sedimentation rate (ESR), white blood cell (WBC) count, and C-reactive protein (CRP) levels. Cultures of blood and biopsy specimens are frequently negative.

Pathophysiology

Bacterial, fungal, and parasitic organisms can cause spinal infections. Staphylococcus aureus accounts for approximately 60% and Enterobacter species about 30% of spinal infections. Other organisms are found most commonly in specific types of patients: Salmonella (in sickle cell disease), Klebsiella and Pseudomonas (in intravenous drug abusers), Streptococcus (in patients with endocarditis and colonic polyposis), and S erratia . S. aureus is known to produce several proteolytic enzymes, including hyaluronidase, which is postulated to cause lysis of the disc. The source for the spinal infection may be concurrent urinary tract, pulmonary, pelvic, or cutaneous infection; contamination from a dirty intravenous needle; or, less frequently, inflammatory bowel diseases, septic abortion, cellulitis, fasciitis, subcutaneous abscess, or pyomyositis. More than 65% of cases of pyogenic spinal infection have an identifiable primary site of infection.

Hematogenous Dissemination

Arterial Route

Hematogenous spread occurs more frequently via the arteries than the veins. The arterial network of ascending arteries, descending arteries, and anastomosing branches, which neighbor the vertebrae, give rise to minute branches that penetrate the cortex and ramify within it. The richest nutrient arteriolar network is located in the subchondral region of the vertebral body, which is the equivalent of the metaphysis of a long bone. Hematogenous organisms arrive in the vertebrae via end-arteriolar arcades in the subchondral plate adjacent to the disc, particularly in the anterior part of the subchondral plate. They disrupt the overlying cortical bone, extend to disc, extend to the opposite vertebral body, and even extend to the subligamentous paravertebral and epidural spaces in adults.

In adults, the intervertebral disc is avascular. However, secondary vascularization may occur with degenerative disc disease, because vascularized granulation tissue grows into the radial tears of older patients. Direct hematogenous spread of infection to the disc may then become possible in these cases. In children younger than the age of 4 years, vascular channels do extend directly into the discs. In these young children, bacteremia may cause direct hematogenous inoculation of the disc with subsequent bacterial discitis.

Venous Route

Valveless veins leave the vertebral body through the central dorsal nutrient foramen and drain into the extradural venous plexus. The extradural venous plexus is interconnected with the valveless paravertebral venous plexus of Batson. Elevated intra-abdominal pressure allows retrograde hematogenous spread of infection from the pelvis and abdominal organs to the vertebral column. The venous route of dissemination to the vertebral bone is of particular importance in infections of the urinary tract and other pelvic organs.

Nonhematogenous Dissemination

Nonhematogenous inoculations are major routes of discitis in adults.

Direct Extension

Nonhematogenous inoculation of the spine may arise secondary to penetrating trauma, direct extension from contiguous infection, or interventional procedures such as biopsy, chemo/mechanical nucleolysis, laser ablations, pain-relieving procedures, sympathectomy, spinal anesthesia, discography, surgical interventions, and spinal instrumentation. Facet injections with corticosteroids for diagnosis or treatment of pain may cause or exacerbate infection of the posterior elements, including the posterior paraspinal soft tissue, the facet joints, the spinous process, and even the pedicles. In the absence of direct iatrogenic inoculation of these elements, however, infection of the posterior spinal elements is uncommon. In the absence of direct inoculation, infection of the laminae and pedicles should raise suspicion of tuberculous spondylitis.

Spread through the Cerebrospinal Fluid

Spinal infection may also arise from extension of infection from the central nervous system (CNS) through the cerebrospinal fluid (CSF) and by CSF leaks.

Pathology

Suppuration may be seen in the paraspinal soft tissues. Meningitis may be present. The infection may spread to, or from, the neural tissue. Suppurative infection may involve the subarachnoid space.

The virulence of an organism may be determined by major bacterial factors such as surface protein receptors, capsular polysaccharides, and toxins. The organism may produce enzymes such as proteolytic enzymes that assist it to invade the disc by lysis of the annulus. With pyogenic bacteria, the initial local response is acute inflammation, producing an exudate containing polymorphonuclear leukocytes (neutrophils) and fibrin. Because bone and disc cannot expand to relieve pressure by swelling, continuing exudation raises the tissue pressure. The result is destruction of bone, with inflammatory exudate, cellular debris, and vascular proliferation.

Imaging

Plain Radiography

The sensitivity and specificity of plain radiographs are very low. They usually lack the ability to show early findings, and negative results do not exclude the presence of infection. Radiographs typically remain normal for 2 to 3 weeks after the onset of infection. The earliest radiographic sign is loss of definition and irregularity of the vertebral end plate, usually starting anterosuperiorly. In the period from 2 to 8 weeks, the intravertebral disc may show an initial increase in height (rarely observed), followed by loss of disc height (seen commonly). Gradually progressive osteolysis causes poor definition of the end plates. Erosions of the cortical end plates on both sides of a narrowed intervertebral disc with associated paraspinal mass are the hallmarks of pyogenic infection.

After approximately 10 weeks, plain radiographs may show reactive sclerosis, new bone formation with osteophytosis, kyphotic deformity, scoliosis, spondylolisthesis, and bony ankylosis. Plain radiographs may also show signs of infection in the soft tissue surrounding the spondylitis, including increase in the retropharyngeal space due to cervical spondylitis, displacement of the parietal pleura due to thoracic spondylitis, and indistinct margins of the psoas muscle due to lumbar spondylitis.

CT

CT has high sensitivity for spondylitis but lacks specificity. It is not the primary modality for diagnosis of early spondylitis and disc space infection or for follow-up of spondylitis. CT may miss epidural involvement due to beam-hardening artifacts, especially in the cervicothoracic region. This limitation can be overcome by performing CT myelograms after the intrathecal injection of a contrast agent. However, performing spinal taps in patients with pyogenic infection may lead to intradural spread of infection. If myelography must be done, the spinal tap should be performed at a site distant from the source of infection, for example, via a lateral C1-C2 tap for the evaluation of lumbar spinal infection.

CT provides excellent depiction of fine bone detail and displays well any associated changes in the paraspinal soft tissue. It is useful for showing osteopenia, soft tissue calcification, cortical bone erosion, permeative bone destruction, lytic fragmentation, bony sclerosis, paraspinal soft tissue infiltration with obliteration of fat planes, compromise of the spinal canal, and gas within the disc, bone, and soft tissue. Hypodensity of the disc and vertebral body are major findings of the infection. Sagittal reformatted and 3D images demonstrate reduced disc height, but spiral (helical) CT with thin slices and multiplanar reconstructions are needed to avoid artifacts from partial volume averaging. CT is very useful for guiding percutaneous biopsies, drainage of fluid collections, irrigation of spaces with antibiotics, and so on ( Fig. 18-1 ). For CT-guided aspirations and biopsies, the target for the needle should be the focus of active abnormal contrast enhancement.

FIGURE 18-1, CT-guided biopsy. CT-guided percutaneous biopsy of a pyogenic spondylodiscitis with extensive bone destruction and a large paravertebral abscess.

MRI

MRI is sensitive, specific, and accurate (96%, 94%, and 92%, respectively), equal to the results of combined nuclear medicine studies. Therefore, MRI is the modality of choice for assessing of possible spondylitis. Early on, infection causes an exudate containing WBC and fibrin within vertebral marrow. The extracellular water content increases, and tissue pressure rises. This inflammatory reaction, consequent ischemia, and reactive bone marrow stimulation appear to be responsible for the abnormal T1 and T2 signal intensity seen with infection ( Fig. 18-2 ). Morphologic alterations, such as loss of definition of the end plates of a single vertebra, become more marked in time ( Fig. 18-3 ). The infection may extend underneath the anterior and posterior longitudinal ligaments or into the disc, leading to signal alteration on precontrast T1-weighted (T1W) and T2-weighted (T2W) MR images and to contrast enhancement on postcontrast T1W images. Initially, infection of the disc space commonly appears as nonanatomic T2 signal in the intervertebral disc, loss of the low signal equatorial band (the “intranuclear cleft”), and reduction of disc height (see Fig. 18-2 ). Thereafter, the infection may cause discontinuity of the adjacent bony end plates and progressive destruction of the vertebral bodies, plus frank disc destruction and soft tissue infiltration ( Figs. 18-3 and 18-4 ). The infection can extend posteriorly into the epidural space and/or laterally into the paraspinal tissue ( Figs. 18-4 and 18-5 ). These extensions can be much better defined on the postcontrast images. Infection may cause engorgement of epidural basivertebral veins by direct extension of the inflammatory process, by mechanical obstruction to venous drainage, or by both. However, simple enhancement of the epidural venous plexus should not be confused with epidural infection.

FIGURE 18-2, Early pyogenic spondylodiscitis. Sagittal T2W ( A ), STIR ( B ), and precontrast ( C ) and postcontrast T1W ( D ) MR images show reduced disc height and low signal of the L4-5 disc. The cortical end plates are not interrupted. At L4-5, the disc and end plates show linear, nonanatomic contrast enhancement. The signal intensity of neighboring vertebral bodies and end plates is low on the T1W images and higher on the T2W and STIR images. No soft tissue involvement is noted. E , Sagittal T2W MR image of a different patient shows reduced disc height with nonanatomic high signal of the disc, loss of the intranuclear cleft seen in the adjacent discs, and infiltration of the inferior end plate of the apposing vertebra.

FIGURE 18-3, Pyogenic spondylodiscitis. Sagittal T2W ( A ) and precontrast ( B ) and postcontrast T1W ( C ) MR images show confluent signal alterations of the L1-5 vertebral bodies. Note the T1 hypointensity and T2 hyperintensity in the portions of the L1-L5 vertebrae adjacent to the intervertebral discs, in the anterosuperior portion of T12, and in the entire L4 vertebra. The end plates show interruption of cortical continuity. There is clear but mild extension to the epidural space. On the sagittal T2W image ( A ) the discs show reduced height, nonanatomic high signal, and loss of the intranuclear cleft. The discs and neighboring vertebral bodies show marked nonanatomic contrast enhancement.

FIGURE 18-4, Spondylitis with epidural phlegmon. Sagittal STIR ( A ), T2W ( B ), precontrast ( C ) and postcontrast T1W ( D ), axial T2W ( E ), and T1W ( F ) MR images show diffuse signal abnormalities in the vertebral bodies adjacent to the L2-3 disc with hypointensity on the T1W images and hyperintensity on the T2W images and STIR sequence. The apposing end plates show erosion and destruction. The disc signal is low on T1W images and high on T2W images. There is marked enhancement of the active infection and mild enhancement of the spondylitis. Homogeneous signal and contrast enhancement within the epidural space strongly suggest concurrent epidural phlegmon. Right-sided paraspinal soft tissue involvement is also noted.

FIGURE 18-5, Late stage spondylitis with paraspinal abscess. Sagittal STIR ( A ), T2W ( B ), precontrast ( C ) and postcontrast T1W ( D ) and axial postcontrast T1W ( E ) MR images demonstrate diffuse signal alteration of the vertebral bodies neighboring L3-4 disc space with hypointensity on T1W images and hyperintensity on T2W and STIR images. Erosion of the end plates with destruction and posterior fusion are evident. Anterior half of the disc signal is low on T1W imaging and high on T2W imaging with absence of the intranuclear cleft. Absence of enhancement of the anterior half of the disc space suggests abscess formation whereas enhancement of adjacent vertebral bodies is seen. Extension of infection to the posterior elements, epidural space, and also paravertebral soft tissue is clearly seen on axial images. The axial postcontrast image demonstrates a paravertebral abscess with ring-like enhancement.

Contrast enhancement increases the conspicuity of the lesion, the specificity of diagnosis, and observer confidence in the diagnosis. Contrast-enhanced studies facilitate treatment planning and monitoring of spinal infection. Contrast enhancement is an early sign of acute inflammation and may persist for weeks or months in subtle infections. Fat-suppression MRI techniques increase lesion conspicuity on contrast-enhanced studies. Measurement of percentage contrast enhancement has been shown to be a reliable method to quantify diffuse bone marrow changes. Measurement of contrast enhancement is also helpful to differentiate among the disc, the body, and the edema, the phlegmon, or the abscess (see Fig. 18-4 ). Such differentiation significantly aids treatment planning because surgical drainage is indicated for an abscess whereas conservative therapy is proper for phlegmon. When serial images show that contrast enhancement no longer occurs, active inflammation can be excluded.

Pseudosparing of the end plates is described as a potential pitfall of MRI diagnosis of spondylitis. If the MR images of the spine are obtained with the frequency encoding in the superoinferior direction, increased conspicuity (“pseudosparing”) of the discs may complicate diagnosis. Pseudosparing can be seen when the normal chemical shift artifact seen in healthy end plates is lost as a result of replacement of the lipid-rich yellow bone marrow with water-based infiltrate. Phase encoding (rather than frequency encoding) is strongly preferred for the craniocaudal direction to reduce normal chemical shift artifact, to prevent pseudosparing, and to reduce pulsation artifacts caused by abdominal vessels.

Diffusion-weighted spinal imaging (DWI) has been shown to help differentiate long-standing infection from Modic type I degeneration and metastases. Apparent diffusion coefficient (ADC) values are more reliable than qualitative evaluations. Further studies with larger patient groups are needed for more accurate conclusions on this subject. Current research is evaluating magnetic resonance spectroscopy (MRS) for the differential diagnosis of the spinal pathology, but MRS is not presently in widespread use.

Vertebrae

Within the vertebral bodies, low signal areas probably result from replacement of fat cells by stimulated proliferating bone marrow cells that form white blood cells (WBCs). These areas are seen more reliably on spin-echo T1W images, which are preferable to gradient-echo T1W images. Bone marrow edema may be more difficult to appreciate on T1W images in children and young patients with red marrow and in patients with Modic type III end plate degeneration.

Infections of the vertebral body and disc typically cause high signal on T2W MR images (see Fig. 18-2 ). In early infectious diseases, this high T2 signal may be obscured by fast spin-echo T2W imaging, owing to the normally bright T2 signal of marrow. Later in the disease, when trabecular sclerosis obliterates the marrow space, and in patients with Modic III degeneration, the decreased T2 signal intensity of sclerosis may mask the increased signal changes of an infectious process. Fat-suppressed T2W and short tau inversion recovery (STIR) techniques increase the conspicuity of infected areas and have slightly higher sensitivity than T2W alone for detecting areas of involvement but also are less specific (see Figs. 18-2 to 18-5 ). In addition, STIR sequences do not depict fine anatomic detail and are sensitive to patient and CSF motion. Bone marrow edema can also be evaluated by using opposed-phase gradient-recalled-echo sequences. Normal marrow exhibits low signal intensity with signal subtraction of the water and fat-bound proton components, whereas edema exhibits high signal intensity. Use of this sequence does not add more information than the routine sequences for the differential diagnosis of a vertebral pathologic process. For pathology of the intervertebral disc, gradient-recalled-echo sequences may provide images of a quality similar to conventional sequences and their use may save time. Contrast-enhanced series show areas of vertebral infection well and are very helpful in the differential diagnosis.

With advanced infection, T2W MR images display end plate erosions as interruptions of the cortical continuity and destruction of the vertebral body (see Figs. 18-3 to 18-5 ). Postcontrast T1W imaging helps to differentiate the disc from the vertebral body. Gas within the vertebral body (intravertebral vacuum clefts) appears as dramatically reduced signal intensity on T1W and T2W images and is pathognomonic of dead bone tissue.

Reactive bone changes, new bone formation, osteophytosis, sclerosis, vertebral body height changes, kyphosis, scoliosis, spondylolisthesis, and ankylosis can all be seen during the late/healing stage of the infective process. Sclerosis is more common in pyogenic spondylitis than in tuberculous spondylitis but is common enough in both that it cannot be used for differential diagnosis.

In infants, spondylitis may present as progressive dissolution of involved vertebral bodies without loss of disc height. Years later, the kyphotic deformity from this infection may mimic congenital kyphosis.

Discs

The normal disc signal and morphology change phasically owing to diurnal variation in disc hydration. These phasic changes must be appreciated and not misinterpreted as a pathologic process. After the second decade, about 94% of normal discs show an equatorial band (intranuclear cleft) of low T2 signal that extends as a line across the equator of the disc. With disc infection, adults typically show reduced disc signal on T1W imaging and increased disc signal on T2W imaging. There is frequently distortion or loss of the normal equatorial band (intranuclear cleft). In children, infection may involve the disc primarily. Acutely, there may be increase in the disc height, followed shortly by decreased disc height (see Fig. 18-2 ). Unlike in adults, infected discs in children show decreased (not increased) signal intensity on T2W imaging. Both children and adults show marked heterogeneous nonanatomic contrast enhancement of the infected disc on postcontrast T1W imaging. Increased signal due to inflammatory disc degeneration should be differentiated from infection. Infected discs show nonanatomic contrast enhancement. Degenerated discs may show “anatomic” peripheral, linear, and/or nodular enhancement due to ingrowth of blood vessels that penetrate the peripheral annulus. In advanced infection the disc is destroyed and cannot be demonstrated. Granulation tissue and osteoid may form new bone that bridges across the annulus. Disc and vertebral changes on T1W images and disc signal alterations on T2W images are reliable findings of infection.

Paraspinal Soft Tissues

Spondylitis is reported to extend to the epidural space in 32% of cases. Routine MRI sequences may not show epidural extension adequately, because the abnormal epidural signal may be isointense with and merge into the high signal of CSF. Proton density–weighted (PDW) and fluid-attenuated inversion recovery (FLAIR) images show higher signal in proteinaceous exudates than in CSF, helping to distinguish the inflamed tissue from CSF. Detection of epidural phlegmon and abscess and differentiation from CSF collections is important for treatment planning. Phlegmon typically enhances as a solid inhomogeneous blush, whereas abscess and necrosis appear as peripherally enhancing masses with a hypointense liquefactive center on postcontrast T1W imaging.

Infected paravertebral soft tissues show varying signal and contrast enhancement (see Fig. 18-5 ). In the acute phase, paraspinal phlegmon appears as ill-defined areas that are hypointense on T1W imaging and hyperintense on T2W imaging, reflecting the extracellular paraspinal edema. There is little mass effect. The necrotic center of a paraspinal abscess is hyperintense on T2W imaging and isointense to hypointense on T1W imaging. The abscess capsule is isointense to hyperintense on T1W imaging and very hypointense on T2W imaging (see Fig. 18-5 ). Contrast enhancement helps to delineate the full extent of soft tissue infection and usually identifies the abscesses as ring-enhancing lesions.

A specific entity—Griesel's syndrome—is characterized by atlantoaxial subluxation, synovial effusion, and inflammation/infection of neighboring soft tissues. Griesel's syndrome is caused by septic emboli transported from infections of the nasopharynx, tonsils, alveoli of the jaw, and lymph nodes to the spine via connections between the pharyngovertebral veins and the periodontoidal venous plexus and suboccipital epidural sinuses. Pyogenic osteomyelitis of the C1 and C2 vertebrae is rare but clinically significant. Cervical instability may result from inflammatory softening or lysis of the bone at the insertions of the transverse ligament. The outcome of the treatment is influenced by the type of infection and by the degree of neurologic compromise before treatment.

The clinical features of spinal infections can be subtle and misleading. Delays in diagnosis can lead to increased morbidity and mortality, so early diagnosis and treatment are essential. However, successful differentiation of infection from degenerative disease, noninfective inflammatory lesions, spinal neoplasm, and other diseases remains difficult, despite advanced techniques for spinal imaging. Imaging criteria sensitive for identifying spondylitis include evidence of paraspinal or epidural inflammatory tissue, contrast enhancement of the disc, erosion and destruction of the vertebral end plates, and abnormal signal on T1W and T2W imaging.

In summary, MRI findings of pyogenic spondylitis and discitis include the following (see also Box 18-1 ):

  • Decreased signal intensity on T1W images and increased signal intensity on T2W images, with abnormal contrast enhancement

  • Irregularity, erosion, and destruction of end plates of vertebral bodies with interruption of the normal signal void of the cortical end plates

  • Reduced disc height, loss of intranuclear cleft on T2W images, disc protrusion, and nonanatomic contrast enhancement

  • Paravertebral soft tissue infiltration/abscess, which appears hyperintense on T2W imaging and hypointense on T1W imaging and shows homogeneous or ring-like enhancement

  • Epidural extension with contrast enhancement (phlegmon: homogeneous; abscess: ring enhancement)

  • Late/healing stage of the infective process: reactive bone changes, new bone formation, osteophytosis, sclerosis, altered height of the vertebral body, kyphosis, scoliosis, spondylolisthesis, ankylosis, and bony bridges across the annulus

BOX 18-1
MRI for Evaluation of Spondylitis

PATIENT HISTORY

A 45-year-old woman with persistent lower back pain of 6 months' duration was admitted to the hospital. She also had malaise, fatigue, and rigidity. Elevated sedimentation rate and white blood cell count were remarkable laboratory findings. Neurologic examination showed radicular symptoms.

TECHNIQUE

MRI examination of lumbar spine with contrast administration was performed. Pre- and postcontrast axial and sagittal T1W, axial and sagittal T2W, and STIR sagittal images were obtained.

FINDINGS

There is normal height and alignment of the vertebral bodies. Confluent change in signal intensity of the L4 vertebral body is seen with hypointensity on T1W imaging, hyperintensity on T2W imaging, and contrast enhancement of the area. The vertebral end plates show interruption of the cortical continuity, erosion, and destruction. Homogeneous signal intensity and contrast enhancement in the epidural space suggest epidural phlegmon. Abnormal soft tissue rim enhancement at the right anterior aspect of the L4 vertebral body suggests soft tissue abscess. Decreased height of the L4-5 disc, nonanatomic high T2 signal within it, loss of its equatorial band (intranuclear cleft), and contrast enhancement are also noted. The differential diagnosis includes pyogenic spondylitis and other spinal inflammatory diseases.

The remaining vertebral marrow signal is normal. The remaining intervertebral discs are normal in height and hydration. There is no evidence of disc herniation, spinal stenosis, or nerve root compression. The spinal cord is normal in size, signal, and caliber.

IMPRESSION

Diagnosis is L4 spondylitis with epidural and paravertebral soft tissue extensions.

There is an increasing move away from open surgical intervention toward conservative therapy, percutaneous abscess drainage, or both. It is critical to monitor treatment response, particularly in the immunodeficient patient. In children, discitis may be treated with antimicrobial therapy and bed rest.

MRI is very helpful for monitoring the success of treatment, because successful therapy is associated with the following:

  • Reduction in paravertebral soft tissue swelling (perhaps the earliest sign of successful therapy)

  • Appearance of a high signal rim at the edge of the lesion (mean, 15 weeks);

  • Higher signal of the involved marrow on T1W and fast spin-echo T2W images than noninvolved marrow (the reconstituted marrow is predominantly fatty and appears to be of higher signal intensity than normal marrow and degeneration of hematopoietic marrow is probably caused by obliteration of the intramedullary vessels, preventing repopulation with red marrow cells)

  • Lower signal within the marrow on both T1W and T2W images (may be a reflection of reactive sclerosis and fibrosis due to healing)

  • Decrease of high marrow signal on STIR, PDW, and T2W images

  • Decrease in the high disc signal on STIR and T2W imaging, even though the disc remains narrowed

  • Reappearance of the equatorial band (intranuclear cleft) (a good indicator of resolution of inflammation)

  • New bone formation bridging the disc space

  • Resolution of canal compromise

  • Progressive reduction in contrast enhancement. (Increasing or persisting contrast enhancement with clinical improvement and increasing destruction does not necessarily indicate treatment failure.)

KEY POINTS: PYOGENIC DISCITIS-OSTEOMYELITIS

  • ▪

    “Good disc bad news, bad disc good news” highlights that the destructive vertebral bone lesion associated with a well-preserved disc space with sharp end plates favors a diagnosis of neoplastic infiltration whereas the destructive bone lesion associated with a poorly defined vertebral bony end plate with or without loss of disc height suggests infection with a better prognosis.

  • ▪

    Exudative pleural effusions may be a manifestation of vertebral osteomyelitis of the thoracic spine. The majority of the effusions have been found to be sterile. Thus, in a patient with pleural effusion of unknown cause, the possibility of vertebral osteomyelitis should be considered, possibly more so in diabetics.

Degenerative End Plate Changes

  • ▪

    Type I degenerative osteoarthritis resembles early pyogenic spondylitis.

  • ▪

    Lack of associated abnormally increased disc signal on T2W imaging (even reduced disc signal), lack of soft tissue involvement, and linear mild contrast enhancement of the disc are the main differentiating findings.

  • ▪

    There may be gas within the severely degenerated discs.

  • ▪

    Schmorl's nodules may be present.

  • ▪

    Cortical continuity of the vertebral end plate is generally intact in the remaining end plate, and there may be signal change of the adjacent vertebral marrow.

Erosive Intervertebral Osteochondrosis

  • ▪

    Inflammatory disc degeneration shows as high signal intensity on T2W imaging. Gas within the disc may also be seen. Focal areas of annular enhancement can be seen with no central enhancement.

  • ▪

    Edema of the vertebral bone marrow adjacent to the disc space uncommonly reaches the middle part and almost never extends to the opposite end plate.

  • ▪

    Erosion of the adjacent end plates occurs with no major destruction.

  • ▪

    No paravertebral/epidural involvement occurs.

  • ▪

    There is minimal or no osteophytosis.

  • ▪

    Dense sclerosis with bone erosions can be seen.

Dialysis Arthropathy

  • ▪

    Decrease of disc height and erosion of the subchondral bone occur at anterior superior and inferior margins.

  • ▪

    Amyloid deposition has a signal intensity similar to that of muscle.

  • ▪

    A clinical history of dialysis for more than 3 years and absence of paravertebral soft tissue infiltration are helpful differential points.

Pseudarthrosis

  • ▪

    Nonunion occurs, resembling destruction, end plate erosions, and sclerosis.

Spinal Metastases

  • ▪

    In the initial stage with no disc space involvement, it is very difficult to differentiate the infection from neoplastic involvement.

  • ▪

    A well-preserved disc space with sharp end plates favors a diagnosis of neoplastic infiltration (in rare instances, metastatic involvement of the disc has been reported).

  • ▪

    Consecutive vertebral involvement is more frequent in spondylitis than in tumoral infiltration.

  • ▪

    The intact vertebral end plate favors the diagnosis of tumor.

  • ▪

    Both infections and tumors may show skip lesions.

  • ▪

    Soft tissue involvement is a diffuse pattern in infections but usually well defined in tumors.

  • ▪

    Signal characteristics were not reliable signs for distinguishing tumors from infection because both of them demonstrated low signal intensity on T1W imaging and high signal intensity on T2W imaging.

Rheumatoid Arthritis

  • ▪

    Signal alterations in the active stage of rheumatoid arthritis can mimic spondylitis.

  • ▪

    Joint effusions, interapophyseal joint erosions, and erosion of the spinous process occur.

Avascular Necrosis

  • ▪

    This is rare.

  • ▪

    Intervertebral vacuum clefts can be seen.

Charcot Spine

  • ▪

    The disease is characterized by destructive process of the vertebral bodies, with end plate and facet joint erosions.

  • ▪

    Joint disorganization is noted.

  • ▪

    Vacuum into the disc and enhancement and decreased disc space are observed.

  • ▪

    There is paraspinal soft tissue swelling.

  • ▪

    Osteophyte formation and sclerosis occur.

Ankylosing Spondylitis

  • ▪

    Vascularized inflammatory tissue of the annulus and longitudinal ligaments shows increased signal on T2W imaging and enhancement.

  • ▪

    Erosions (Romanus lesions) and even destruction of entire discovertebral junction occur in advanced cases.

  • ▪

    End plate and vertebral enhancement can be seen adjacent to the disc space and ligaments.

  • ▪

    Associated sclerosis (subchondral bone formation) neighboring the erosions, syndesmophytes, apophyseal joint fusions, and osseous fusions occur.

  • ▪

    Calcification and ossification of the thickened ligaments occur.

  • ▪

    Sacroiliitis is pathognomonic.

  • ▪

    Disc bulging can also be seen.

Chronic Recurrent Multifocal Osteomyelitis

  • ▪

    This is characterized by relapses and remissions that are unresponsive to treatment; it mimics spondylitis.

Neuropathic Arthropathy

  • ▪

    Disc bulging, disorganization, dislocation, destruction of the vertebras, and spondylolisthesis are seen.

  • ▪

    Soft tissue masses and osseous debris are noted.

  • ▪

    There is rim enhancement of the discs and diffuse enhancement of the vertebral body.

  • ▪

    Spinal cord injury and syrinx are associated.

Lymphoma

Multiple-Solitary Myeloma

PYOGENIC EPIDURAL/SUBDURAL ABSCESS

This is defined as infection of the spinal extradural and subdural spaces with abscess formation. It is also called spinal epidural/subdural empyema.

Epidemiology

Once considered rare, spinal epidural empyema is now seen with increasing prevalence. Spinal subdural infection remains very rare compared with spinal epidural abscess (SEA), with fewer than 50 cases reported in the literature. Diabetes mellitus, intravenous drug abuse, chronic renal failure, excessive alcohol ingestion, immunodeficiency, and local steroid injections are major risk factors for developing these infections. No genetic or racial predilection has been identified. SEA is seen at all ages but is most frequent in the sixth and seventh decades. Patients may show acute deterioration due to mechanical compression or ischemic compromise of the spinal cord, prior local administration of corticosteroids for pain control, or direct infection of the spinal cord. Surgery is indicated when there is neural/cord compression, progressive neurologic deficit, or persistent severe pain.

Clinical Presentation

Clinical diagnosis is challenging. Symptoms of acute infection include back pain, fever, or both, with or without neurologic (mostly sensory) deficits. The precise nature of any neurologic signs depends on the level of infection. Chronic infections, especially tuberculosis, may not manifest as fever or other clinical signs of infection. Instead, patients with chronic infection present widely varying symptoms ranging from mild back pain to paraplegia. A progressive form of subdural empyema shows sudden onset with rapid progression over 2 to 5 days, leading to marked paralysis, loss of bowel and bladder control, and respiratory distress secondary to denervation of the intercostal muscles and diaphragm. Without treatment, patients with SEA usually show a progressive course from pain to weakness of voluntary muscles and sphincters to paralysis.

Pathophysiology

The major routes of infection are hematogenous spread from extraspinal sources (most common), direct inoculation, and contiguous spread. S. aureus is the most common agent (57%-73% of reported cases). The sites most frequently involved are the lower thoracic and lumbar regions. Spinal epidural collections are almost always located at the posterior aspect of the spinal canal. Anterior collections are most frequent with vertebral osteomyelitis. SEA associated with spondylitis typically extends over a limited number of adjoining segments. SEA not associated with vertebral osteomyelitis tends to be more extensive and multifocal. Very long segment SEA often spirals around the spinal canal to lie posterior in some segments and anterior in others.

Pathology

Osteomyelitis may be observed. Meningitis and myelitis may be associated. Frank pus is located in the subdural or epidural space. Leukocytes, cellular debris, vascular proliferation, and granulation tissue are detected on histologic study.

Imaging

Radiography

Plain radiographs provide little benefit for early detection of spinal epidural infection. When spondylodiscitis is associated, it may be evident.

CT

CT may help to demonstrate epidural and paraspinal extensions and associated bone lesions in detail. Myelography with an intrathecal contrast agent may delineate the site(s) of the collections and their relationship to the spinal cord, but the spinal tap needed to perform myelography may spread the infection to previously uninfected sites and is not recommended when infection is suggested.

MRI

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here

Related Posts