Approach Considerations
The Infectious Diseases Society of America (IDSA) has published clinical practice guidelines for the diagnosis of native vertebral osteomyelitis (NVO) in adults (see Guidelines). [59]
Although magnetic resonance imaging (MRI) is the imaging method of choice for vertebral osteomyelitis and diskitis in the early stages, it may show only subtle, nonspecific endplate subchondral changes; a repeat examination may be required to demonstrate the typical features.
Computed tomography (CT) is useful for assessing bony destruction and instability in patient with destructive lesions. Flexion-extension radiographs are also useful for diagnosing biomechanical instability.
Blood tests (eg, complete blood count [CBC], erythrocyte sedimentation rate [ESR], and C-reactive protein [CRP] level) are nonspecific markers of inflammation. The gold standard diagnostic test is a biopsy with tissue cultures for microbiology. [60, 61, 62]
Laboratory Studies
Leukocytosis, the usual indication of infection, is often absent or minimal in patients with chronic pyogenic vertebral osteomyelitis.
Elevation of the ESR, though nonspecific, is the most common laboratory abnormality. Back pain coupled with an increased ESR should lead the clinician to suspect vertebral disease (eg, infection, neoplasia, or rheumatoid disorder).
Blood cultures should always be obtained before administration of antibiotics.
CRP, synthesized by hepatocytes, is an excellent indicator of inflammation. Patients with bacterial diskitis have higher serum CRP and fibrin levels. Patients with nonseptic diskitis (ie, chemical diskitis) have only dense fibrotic histologic changes, and serum CRP and fibrin findings are normal.
The use of soluble urokinase-type plasminogen activator receptor (suPAR) has been proposed as a means of distinguishing vertebral osteomyelitis from degenerative diseases of the spine. In a study by Scharrenberg et al, suPAR was found to be less sensitive but more specific than CRP for this purpose, suggesting that diagnostic power might be enhanced by combining the two. [63]
Plain Radiography
The process of diagnosing a spinal infection usually begins with a radiograph, though radiographic findings are usually normal in the first 2-4 weeks. If the disk space is involved (diskitis), the disk space may narrow, and destruction of the endplates around the disk may be seen on the radiograph. (See the image below.)
Spinal infections. Lateral plain radiographs of Patient A with diskitis at C4-5. Note severe disk space narrowing and subluxation seen at C4-5.
Later, plain radiographs usually reveal rarefaction, loss of bony trabeculation close to the cartilaginous plate, and irregular narrowing of the vertebral disk space. Vertebral body collapse may also be seen (see the image below). Simultaneously, evidence of rapid bone regeneration may be evident, with the development of bone spurs and dense new bone. A paravertebral soft-tissue mass may also be present.
Spinal infections. Patient B (47-year-old woman) presented with intractable back pain. Radiographs reveal significant collapse and destruction of the L4 vertebral body. MRI of lumbar spine was ordered.
CT and PET
CT depicts osteomyelitis earlier than plain films do. CT findings include hypodensity at the site of infected disks, lytic fragmentation of the involved bone, gas within an involved vertebra, and decreased density of adjacent vertebrae and nearby soft tissues. Epidural and paraspinal extension of infection may also be seen.
The use of 18F-fluorodeoxyglucose (FDG) positron emission tomography (PET) in conjunction with CT is helpful in the diagnosis of vertebral osteomyelitis. In a study by Kouijzer et al, which compared 18F-FDG-PET/CT and MRI with the clinical diagnosis, 18F-FDG-PET/CT had a sensitivity of 100%, a specificity of 83.3%, a positive predictive value of 90.9%, and a negative predictive value of 100% for diagnosing vertebral osteomyelitis. [64] In particular, 18F-FDG-PET/CT had an advantage for visualization of metastatic infection, especially in bacteremic patients.
A study of 133 cases of vertebral osteomyelitis by Russo et al found FDG-PET/CT to be more reliable than MRI for follow-up of infection. [65]
Magnetic Resonance Imaging
MRI of the spine provides information that CT does not. [66, 67] Characteristic MRI findings include destructive and expansile lesions involving two adjacent vertebrae and their intervening disk.
Low-density changes in bone and disk are seen on T1-weighted images, whereas high-density changes are seen in these structures on T2-weighted images, presumably from their increased water content. Intravenous (IV) infusion of gadolinium shows enhancement of the involved structures. Paravertebral infection, collections under the posterior longitudinal ligament, and epidural abscesses may also be shown. (See the images below.)
MRI of Patient B reveals enhancing mass affecting L4 vertebral body with compromise of spinal canal. Patient underwent several blood cultures and CT-guided trocar biopsy; culture results were negative. Surgical procedure was necessary.
Spinal infections. T2-weighted MRI of Patient A. Evidence of osteomyelitis and diskitis, as well as small epidural abscess, is present. The patient underwent C4-5 anterior cervical diskectomy and arthrodesis using autologous iliac crest bone graft and instrumental fixation with titanium plate and screws.
Diffusion-weighted imaging is useful in distinguishing between degenerative and infectious endplate abnormalities. [68] Compared with PET, diffusion-weighted MRI costs less, has faster imaging times, and does not involve the use of ionizing radiation. [69]
Radionuclide Scanning
Radionuclide scans with technetium-99m are very sensitive early indicators of pyogenic vertebral osteomyelitis. Radionuclide scan findings become positive long before plain film changes are evident.
Technetium-99m bone scanning is not useful for specifically differentiating infection from metastasis or osteoarthritis. Gallium is more likely to localize an inflammatory lesion, and technetium-99m combined with gallium-167 demonstrates virtually all pyogenic vertebral infections. [70]
Myelography
In the past, myelography was used in the evaluation of vertebral osteomyelitis to delineate areas of epidural spread and neural compression. MRI has largely supplanted myelography because of its ability to depict not only bony changes but also pus and granulation tissue under the posterior longitudinal ligament and epidural infection.
Other Tests
Patients with vertebral osteomyelitis can develop urinary retention. Urodynamic studies may be helpful. Methods of objectively testing the behavior of the lower urinary tract during filling, storage, and micturition include the following:
-
Uroflowmetry
-
Cystometry
-
Sphincteric electromyography (EMG)
-
Combined studies
When appropriately used, urodynamic testing provides valuable information for the evaluation and subsequent treatment of neurourologic dysfunction.
Procedures
CT-guided percutaneous biopsy of the infected vertebra or disk may be done via a needle or trocar. Findings are positive only 60-70% of the time. This is a minimally invasive test used to obtain histologic confirmation of the disease and tissue samples for culture. Trocar biopsies have proved more useful than fine-needle aspiration (FNA) because they allow a larger amount of material from the infected area to be examined histologically, as well as cultured. Fluoroscopy may be an option for biopsy guidance if CT is not available. [71]
As with blood cultures, the likelihood of positive tissue culture findings decreases if antibiotic therapy has already been initiated. A 10-year retrospective review suggested that paravertebral soft tissues may also be considered viable biopsy targets. [72]
If blood cultures and percutaneous biopsy fail to identify the infecting organism, open surgical biopsy is indicated. An open surgical biopsy has the highest yield in terms of positive culture findings and diagnostic confirmation. [73]
Histologic Findings
Histologic findings are similar to those of any bacterial pyogenic infection. Local destruction of the disk and endplates occurs with infiltration of neutrophils in the early stages. Later, a lymphocytic infiltrate predominates.
-
Spinal infections. Lateral plain radiographs of Patient A with diskitis at C4-5. Note severe disk space narrowing and subluxation seen at C4-5.
-
Spinal infections. T2-weighted MRI of Patient A. Evidence of osteomyelitis and diskitis, as well as small epidural abscess, is present. The patient underwent C4-5 anterior cervical diskectomy and arthrodesis using autologous iliac crest bone graft and instrumental fixation with titanium plate and screws.
-
Spinal infections. Patient B (47-year-old woman) presented with intractable back pain. Radiographs reveal significant collapse and destruction of the L4 vertebral body. MRI of lumbar spine was ordered.
-
MRI of Patient B reveals enhancing mass affecting L4 vertebral body with compromise of spinal canal. Patient underwent several blood cultures and CT-guided trocar biopsy; culture results were negative. Surgical procedure was necessary.
-
Spinal infections. Patient B developed lower-extremity weakness, and follow-up studies reveal further compression of L4 and compromise of canal. Anterolateral approach was performed with corpectomy, decompression of spinal canal, restoration of anterior column support, and arthrodesis with titanium cage and autologous iliac crest bone graft. Pathology and Gram stain revealed some hyphae. Culture findings were positive for Aspergillus species. Patient underwent full course of amphotericin B and completely recovered.
-
MRI with gadolinium contrast in 41-year-old patient with history of IV drug use and extensive osteomyelitis, diskitis, and anterior spinal epidural abscess causing compression of spinal cord.
-
Intraoperative fluoroscopy in 41-year-old patient with history of IV drug use and extensive osteomyelitis, diskitis, and anterior spinal epidural abscess treated with corpectomy and anterior stabilization.
-
AP intraoperative fluoroscopy. Given abnormal softening of bones by osteomyelitis, same patient underwent posterior stabilization of spine with lateral mass and pedicle screws.
-
3D CT of 41-year-old man with severe osteomyelitis and diskitis at L3 and L4. Note bony destruction. This was later demonstrated with gadolinium-enhanced MRI.
-
Gadolinium-enhanced MRI in 37-year-old male with history of IV methamphetamine use who developed severe diskitis at L3-4. Epidural abscess is also present. Treatment of his condition required extensive surgical decompression and stabilization. Cultures showed polymicrobial infection, including gram-positive and gram-negative bacteria. Antibiotic beads soaked in vancomycin and tobramycin were applied to surgical site.
-
Gadolinium-enhanced MRI in 67-year-old male with L3-4 and L4-5 diskitis. Initial cultures were negative. He was treated medically with 8 weeks of IV antibiotics and clinical follow-up with ESR and CRP.
-
Same patient presented several months later with intractable pain. His CT and MRI showed extensive destruction of disks and vertebral bodies. His condition required extensive debridement with stabilization of spine using transpedicular screws and rods. Antibiotic beads were placed at surgical site.
