Incidence and Demographics

Spontaneous diskitis has an incidence of 0.2 to 2.4 per 100,000, with a bimodal distribution that peaks in early childhood and again at around 60 to 70 years of age.1,2 The overall incidence of spontaneous vertebral osteomyelitis is approximately 2.4 per 100,000, with an incidence that increases with age. The incidence is 0.3 per 100,000 in those younger than 20 years of age and 6.5 per 100,000 in those older than 70 years of age.3 Vertebral column infection has a male predominance.3–5 Patient risk factors for spontaneous diskitis include intravenous drug use, alcoholism, diabetes mellitus, malignancy, immunosuppression, human immunodeficiency virus (HIV) infection, endocarditis, renal failure, and cirrhosis.1,2,6 Similar risk factors are associated with spontaneous vertebral osteomyelitis.5,7,8

Some fungal and atypical bacterial infections are endemic in certain areas. Blastomycosis is more common around the Great Lakes and Ohio and Mississippi River basins. Coccidioidomycosis is localized to the southwestern United States and Central America.9 In countries around the Mediterranean Sea, brucellosis comprises a sizable portion of vertebral column infections. Spinal tuberculosis, also known as Pott disease, is more common in Southeast Asia and Africa ( Fig. 12.1 ).10

The overall rate of infection after spine surgery is 2.1% (range, 0.5 to 12%), with the procedure type a main determinant of the postoperative infection rate.11–20 In general, longer procedures and operations with instrumentation have higher rates of infection. The incidence of infection after procedures involving intervertebral disks is relatively low, at 0.5 to 1%.21,22 Simple laminectomy without bony fusion has an infection rate of approximately 1.5 to 2%.22,23 With the addition of instrumentation and grafting, the rate of infection increases significantly to 2.8 to 6%.22–25 The infection rate is increased still further if the surgery is done in the setting of spine trauma, with a reported risk of 10%.26

A large number of patient characteristics have been associated with an increased risk for spine surgical site infections: obesity, malnutrition, increased age, steroid use, tobacco abuse, alcoholism, diabetes mellitus, malignancy, concomitant urinary tract infection, prior surgery, prior infection, prior radiation therapy, prolonged preoperative hospitalization, trauma, and complete neurologic deficit before surgery.20,23,24,26–35 The presence of three or more comorbidities further increases the risk for postoperative infection.20

Etiology and Pathogenesis

Infections of vertebral column are most often the result of direct implantation during invasive procedures and surgery. A much less common source of direct implantation is penetrating trauma.36,37 Spontaneous infections are usually caused by hematogenous spread from a remote infection, such as endocarditis, dental abscess, skin infection, or urinary tract infection. This hematogenous spread can be either arterial or venous. The paravertebral venous plexus (Batson plexus) is a valveless venous system that connects the internal vertebral plexus to the deep pelvic and thoracic veins, thus allowing a conduit for the distant spread of infection to the vertebral column.38 Finally, infection may spread to the vertebral column from contiguous infected structures, such as overlying infected skin and decubitus ulcers, as well as deep infections, such as pyelonephritis, mediastinitis, and pharyngeal abscesses.

The bimodal distribution of ages seen in patients with spontaneous diskitis is likely due to the differences in vascular anatomy between children and adults. During childhood, end-arteries penetrate the nucleus pulposus of the intervertebral disk, providing a direct channel through which bacteria can enter the disk. Also, the vertebral end plates have more arterial anastomoses in children, allowing better bacterial clearance. By adulthood, the arterial supply reaches only the annulus fibrosus, decreasing the likelihood of direct bacterial embolization into the disk, whereas involvement of the vertebral end plates increases. The rate of diskitis increases again in late adulthood because of the increase of comorbidities with age.1,39

Staphylococcus aureus is the most common organism to cause diskitis, followed by gram-negative rods, Streptococcus species, and Enterococcus species.2,4 The likelihood of gram-negative rod diskitis increases in patients with gastrointestinal or genitourinary infection, diabetes mellitus, intravenous drug use, or a suppressed immune system.6 Fungal diskitis is uncommon but is more likely in patient populations that are immunocompromised, neutropenic, critically ill, on multiple antibiotics, or that have indwelling central venous catheters.40 Atypical bacteria, such as those causing tuberculosis or brucellosis, can involve the disk space but are much more likely to involve the vertebrae. Patients with sickle cell disease have an increased risk for infection with Salmonella species and Haemophilus influenzae.6

Spontaneous vertebral osteomyelitis generally begins at the vascular vertebral body end plates; therefore, the vertebral body is almost always involved, whereas the posterior elements are only rarely involved (3 to 12%).41 The most common source for spontaneous vertebral osteomyelitis is urinary tract infection, followed by skin infection.5

As with diskitis, the most frequent organism causing vertebral osteomyelitis is S. aureus, followed by gram-negative rod species, especially Escherichia coli.5 Pseudomonas aeruginosa is relatively more common in intravenous drug users, but even in this patient population S. aureus is the most frequent causative organism.7 Tuberculosis remains an important cause of osteomyelitis in many developing regions of the world and in certain at-risk populations, such as patients with HIV infection. Brucellosis is another atypical bacterial cause of osteomyelitis that is found in the Mediterranean region and is associated with animal contact or the ingestion of unpasteurized milk.42 Other causes of spontaneous vertebral osteomyelitis include various fungal infections, such as candidiasis, aspergillosis, coccidioidomycosis, histoplasmosis, and blastomycosis, and rarely parasitic infections such as echinococcosis.9,40,43

The distribution of the involved spinal segments is similar for both diskitis and osteomyelitis, with the lumbar spine most frequently involved (60%), followed by the thoracic spine (30%) and the cervical spine (10%).1,4,5 Usually, one spinal level is involved, but multilevel involvement is seen in 5 to 18% of pyogenic cases and 20% of tuberculosis cases.1 Vertebral osteomyelitis is frequently associated with adjacent paravertebral abscesses (26%) and epidural abscesses (17%).8

Postoperative vertebral column infections are caused by the direct spread of bacteria into the surgical wound from surgical instruments, the surgeon′s hands, or airborne transmission. For early-onset infections, the most common organism is S. aureus, which accounts for 50 to 75% of cases, followed by Staphylococcus epidermidis, gram-negative rod species, and polymicrobial infections.22,23 S. epidermidis, Propionibacterium acnes, and Corynebacterium species are often the cause of late-onset infections associated with instrumentation.44

Decreasing exposure of the surgical wound to bacterial contamination aids in preventing postoperative infection. Adequate cleansing of the skin before the start of surgery is vitally important. Randomized controlled trials have demonstrated that chlorhexidine-alcohol is superior to povidone-iodine for preoperative skin cleansing.45 Likewise, chlorhexidine-based scrubs and alcohol rubs are superior to povidoneiodine–based scrubs for decolonizing hands before surgery.46 The practice of double gloving reduces the incidence of glove perforations that can result in cross-contamination.47 Vertical laminar airflow decreases the exposure of surgical wounds to airborne contamination and decreases the rate of infection in posterior spine fusions.48 Frequent irrigation decreases the bacterial burden in direct contact with the wound.

During any surgery there will be some bacterial contamination because perfect sterility is not possible. Any factor that prolongs surgery will theoretically increase exposure to contamination and increase the risk for infection. Operative times longer than 3 hours increase postoperative infection rates.49 Multilevel surgical fusions also increase the risk for infection, in part because of the prolonged operative time, although the addition of a foreign body likely also plays a role.50

The ability of the body to defend itself against infection also plays an important role in determining whether a postoperative infection will develop. This is governed in large part by underlying patient conditions, as described previously. Surgical factors that influence the immune response are also critical. A large volume of blood loss that is greater than 1 L or requires blood transfusion has been associated with increased rates of infection.24,32 Prolonged retraction can devitalize tissue and provide an environment conducive for bacterial growth. This may explain in part why posterior spinal surgeries, which require the extensive use of retractors, have a significantly higher rate of infection than anterior spinal surgeries.26,51 Retraction should be periodically relaxed during prolonged surgery. The gentle handling of tissues and meticulous reapproximation of skin edges will also aid in healing.

Finally, perioperative intravenous prophylactic antibiotics can decrease bacterial viability and prevent postoperative infections.52,53 The appropriate use of prophylactic antibiotics decreases the rate of postoperative spine infections from 5.9 to 2.2%.54 Prophylactic antibiotics should be given just before incision, should be dosed again during long procedures, and should be continued for no more than 24 hours. Topical prophylactic antibiotics applied to the wound bed may also help prevent postoperative infections. Gentamicin-soaked collagen sponges placed in the disk space may decrease the rate of postoperative diskitis, and vancomycin powder applied to the surgical wound before closure has been shown to decrease infections in instrumented spine fusions; however, irrigation with antibiotic-containing saline has not been proven to have additional benefit over irrigation with normal saline alone.52,55–58

Clinical Presentation

The major presenting symptom of both spontaneous diskitis and osteomyelitis is localized pain, with approximately 90% of patients reporting back or neck pain, depending on the region involved.1,8 The pain caused by a vertebral column infection is less likely to decrease with rest or recumbency than back or neck pain caused by degenerative conditions and is often more resistant to analgesic medications. Radicular pain related to irritation or compression of adjacent nerve roots can occasionally occur. Patients with lumbar involvement may have pain elicited with a straight leg raise test. The pain associated with diskitis in children can present as a refusal to walk or bear weight.39 Osteomyelitis may also present with a sudden, severe worsening of pain associated with vertebral body fracture. Localized tenderness to percussion or palpation occurs in approximately one-fifth of patients with osteomyelitis.59

Another common symptom of spontaneous vertebral column infection is fever. The association of back or neck pain with fever should always raise the possibility of spine infection; however, fever is not universally present. Only 60 to 70% of patients with diskitis and 35 to 60% of those with osteomyelitis report fever.2,8,59 Other, less frequent systemic findings include chills, night sweats, general malaise, and anorexia.59

Neurologic deficits, including weakness, difficulty walking, sensory abnormality, urinary incontinence or retention, fecal incontinence, and reflex abnormality, can all occur and are reported in approximately one-third of osteomyelitis cases.59 Neurologic deficits are less likely in diskitis. Paralysis or rapidly progressing deficits are concerning for concurrent spinal epidural abscess. Chronic osteomyelitis may present as a worsening kyphosis, most pronounced in the sharply angled gibbous deformity, which can also cause spinal cord compression and neurologic deficits.8 Chronic vertebral osteomyelitis may present with a draining sinus tract.10

Vertebral column infections are often initially misdiagnosed because the most common presenting symptom, pain, is nonspecific. The mean time to diagnosis for vertebral osteomyelitis is 1.8 months. In one series, the diagnosis of vertebral osteomyelitis was considered in only a quarter of patients at initial evaluation.8 The differential diagnosis of vertebral column infection includes infections of the spinal canal, vertebral column fracture, degenerative disk disease or herniated disk, metastatic or primary malignancy, inflammatory spondyloarthopathies, neuropathic arthropathy, sarcoidosis, psoas abscess, pancreatitis, and pyelonephritis.60

Worsening pain is also a frequent symptom in postoperative vertebral column infections. Differentiation between the expected pain of surgery and the pain caused by infection can be difficult. Pain that returns after initial postoperative relief of preoperative symptoms and worsening localized pain remote from the time of surgery are concerning for infection. Diskitis can present as just worsening pain, more than a month after diskectomy, in a patient with a well-healed incision and no fevers. Recurrent disk herniation is sometimes mistakenly diagnosed, or the patient may be wrongly accused of malingering. An evaluation for infection should be considered in patients with recent spine surgery and worsening pain.

Fever is frequently, but not always, present in postoperative vertebral column infections. Moreover, several other common causes for fever postoperatively must be considered, including atelectasis, pneumonia, urinary tract infection, drug reaction, and deep vein thrombosis.

Wound drainage is the most common sign of postoperative spine infections and can be the only sign or symptom in some patients. Approximately 93% of patients with postoperative spine infection have drainage. Other external signs of infection include erythema, swelling, tenderness, and wound dehiscence. The time from surgery to initial presentation of a postoperative spine infection is around 2 weeks. Infection with aggressive organisms like Clostridium perfringens, however, can present in days, and infection with indolent organisms like P. acnes can present years after surgery.22

The development of new neurologic deficits not initially present postoperatively should warrant urgent evaluation to rule out abscess causing spinal cord or nerve root compression. Treatment of an abscess causing spinal cord compression is a neurosurgical emergency because delay in treatment may dramatically increase morbidity or mortality.61

Diagnosis

Imaging

Plain radiographs are not particularly sensitive or specific for the diagnosis of vertebral column infections, and findings may be normal for weeks after the onset of infection. Paravertebral soft tissue swelling, which is more noticeable in the cervical region, can sometimes be the first radiographic sign of vertebral column infection. Disk space narrowing generally takes between 7 and 10 days to develop but is common in degenerative spine disease as well. Vertebral end plate erosion and vertebral body rarefaction take weeks to develop. Vertebral fracture or collapse can develop late and is sometimes confused with osteoporotic fracture. In patients with kyphotic deformity in the setting of vertebral osteomyelitis, the comparison of recumbent and standing radiographs can aid in the determination of overt instability. In general, plain radiographs may be useful as an initial screening study and can also be used as a follow-up study to monitor for new or worsening kyphotic deformity ( Fig. 12.2e,f ).

CT information is similar to that of plain radiographs but is more sensitive to early changes and has markedly increased anatomic detail. CT is the best study for evaluating the features of bony involvement and is vital for preoperative planning in vertebral column infections requiring surgical fixation ( Fig. 12.2a ). Some soft tissue details are also discernible on CT, such as the presence of paravertebral fluid collections. The administration of intravenous iodinated contrast will usually result in the enhancement of infected disk and bone as well as rim enhancement of associated abscesses. Contrasted CT is particularly useful for the imaging guidance of percutaneous biopsies or abscess drainage. Although contrasted CT is not as sensitive or specific for vertebral column infections as contrasted magnetic resonance (MR) imaging, it can often provide adequate diagnostic images in patients who are unable to undergo MR imaging because of implanted devices or morbid obesity.68

Myelography is generally avoided because of the availability of adequate noninvasive techniques and the risk for seeding the cerebrospinal fluid by performing a lumbar puncture in the setting of infection.

MR imaging combined with gadolinium contrast administration is the most reliable imaging modality for diagnosing spontaneous vertebral column infections. On noncontrasted T1-weighted images, diskitis and osteomyelitis will demonstrate hypointensity of the disk and vertebral bodies, as opposed to normal hyperintensity in the vertebral bone marrow due to the presence of fat ( Fig. 12.2b ). Intravenous gadolinium will enhance regions of inflammation, especially the vertebral end plates and associated abscesses ( Fig. 12.2c ).68 The involved bone and disk will usually be hyperintense on T2-weighted images because of increased edema ( Fig. 12.2d ). With diskitis, there is usually a loss of the intranuclear cleft and decreased disk space height.69 End plate destruction, frank vertebral collapse, and kyphotic deformities are late findings of vertebral column infections. With isolated diskitis, MR imaging has a sensitivity and specificity of over 90%, and with vertebral osteomyelitis, the sensitivity is 96% and the specificity is 93%.68,70 Imaging the entire spine is recommended by some authors because of the possibility of concomitant distant vertebral column infection.71

Radionuclide studies are sometimes useful when MR imaging is either not possible or nondiagnostic, such as in the setting of hardware-induced artifact.68 They rely on increased uptake of various radioactive tracers in metabolically active regions, such as infection, but will also label inflammation in general. They are less sensitive than MR imaging but more sensitive than contrasted CT. Gallium-67 singlephoton emission computed tomography (SPECT) and combined gallium-67/technetium-99 (bone) SPECT have the highest reported sensitivity (91%) and specificity (92%). SPECT scans are more accurate than scintigraphy.54 Labeled leukocyte imaging, although useful in the diagnosis of osteomyelitis elsewhere in the body, has been found to be less effective for the diagnosis of vertebral osteomyelitis. [¹⁸F]Fluorodeoxyglucose positron emission tomography (FDG-PET) is another modality that has, in preliminary studies, identified vertebral osteomyelitis with good accuracy.72

Postoperative vertebral column infections have many of the same radiologic characteristics mentioned above, but the interpretation of imaging is often made more difficult because surgery itself causes inflammatory changes that can mimic infection.73 Contrasted CT and MR imaging can both identify enhancing fluid collections concerning for abscess, although care must be taken not to confuse abscesses with other postoperative fluid collections, such as hematomas and seromas. A plain radiograph can be useful for ruling out retained foreign body. Both radiographs and CT scans can demonstrate the loosening of implants, which can be associated with late-onset infections.74

Related posts:

Epidural and Subdural Infections Meningeal Infections Spinal Canal Infections Infectious Intracranial Aneurysms Spinal Canal Infections Radiology of Central Nervous System Infections

Stay updated, free articles. Join our Telegram channel

Join