Epidural Abscess

An epidural abscess is a rare but potentially life-threatening disease that requires early detection and prompt management. It is defined as an inflammation that involves a collection of pus between the dura (the outer membrane that covers the brain and spinal cord) and the bones of the skull or spine. Spinal epidural abscess (SEA) and intracranial epidural abscess (IEA) are the two types of epidural abscess, and the difference is based on where they develop within the CNS and some variations in risk factors (see Pathophysiology) and symptoms (see History).

A loose association between the dura and vertebral bodies enables extension of spinal epidural abscess to numerous levels, frequently resulting in extensive neurological findings and often necessitating multiple laminectomies. The lumbar and thoracic spine are more commonly affected than the cervical spine.

Tight adherence of the dura to the skull limits expansion of intracranial epidural abscess, often resulting in dangerously increased intracranial pressure, which is a neurosurgical emergency.

Early recognition of these diseases and timely consultation with a neurosurgeon and infectious disease specialist is vital to optimizing the neurological outcome.

Spinal epidural abscess

Causes of spinal epidural abscess

Ten to thirty percent of spinal epidural abscesses result from direct extension of local infection, usually vertebral osteomyelitis, psoas abscess, or contiguous soft-tissue infection.[1, 2, 3, 4, 5, 6, 7]

About half are due to hematogenous seeding. The most likely source is a soft-tissue process, but anything capable of causing bacteremia can result in spinal epidural abscess (endocarditis, urinary tract infection, respiratory tract infections, intravenous drug use, vascular access devices). Hematogenous seeding of the spinal epidural abscess can result in multilevel noncontiguous spinal epidural abscess.

Fifteen to twenty-two percent of spinal epidural abscesses are due to invasive procedures or instrumentation.[8] Spinal surgery, epidural anesthesia, steroid and pain-relieving injections, and placement of pain pumps are all associated with spinal epidural abscess. Short-term epidural anesthesia is much less risky than a catheter left in place for days or permanently implanted. Rates of infection after intraoperative epidural block are about 1 in 2,000, while longer-duration (days) epidural pain catheter placement may be associated with rates of infection as high as 4.3%. Simple epidural injections rarely cause infection; the risk is about 1 in 10,000 injections.[9]

In some cases (up to 30% in some series), the source of the spinal epidural abscess is not identified.

Risk factors for spinal epidural abscess

The most common risk factor for spinal epidural abscess is diabetes mellitus, followed by spinal trauma (may be remote) or surgery, intravenous drug abuse, alcoholism, renal insufficiency, immunosuppression (including infection, steroid use, cirrhosis, and malignancy), central lines, implantable devices (eg, pacemakers), pregnancy, and spinal/epidural anesthesia or injections.[1, 2, 7, 10, 11]

Intravenous drug use is an increasing risk factor in many series.

Anatomy of spinal epidural abscess

Most abscesses occur posteriorly. An anterior location is often associated with vertebral osteomyelitis or a psoas abscess.[1, 2]

The thoracic and lumbar areas are the most likely sites of involvement, with the cervical spine accounting for approximately 20% of cases.[5]

Spread to multiple vertebral levels is common and occurs as the abscess extends up and down the spinal dural sheath. In some cases, this process involves most or all of the spine.

Mechanism of injury [8]

Direct compression of the cord is clearly a major factor.[1]

Vascular occlusion due septic thrombophlebitis and/or vasculitis is also a factor

Infarction of the cord due to loss of arterial blood flow 

The exact mechanism of injury remains controversial.

Intracranial epidural abscess

Because intracranial epidural abscess can cross the cranial dura along emissary veins, an accompanying subdural empyema is often present.[6]

Risk factors for intracranial epidural abscess include prior craniotomy, head injury, sinusitis, otitis media, and mastoiditis.[12, 13]

Frequency

United States

The annual incidence of spinal epidural abscess has risen in the past 2-3 decades from 0.2-1 cases per 10,000 hospital admissions to 2.5-5.1 per 10,000 admissions.[1, 7] The rising incidence of spinal epidural abscess has been attributed to the increasing prevalence of injection drug use, as well as to an increased performance of invasive spinal procedures .Better and more available diagnostics (MRI) may also be contributing to the rising numbers of cases.

The annual incidence of intracranial epidural abscess is difficult to determine but is recognized to be much less common than spinal epidural abscess.

International

Few data on epidural abscesses are available outside the United States, but the frequency appears to be similar to that in the United States.

Mortality/Morbidity

Spinal epidural abscess: At the beginning of the 20th century, almost all individuals with spinal epidural abscess died. However, associated mortality rates have dropped significantly over the past 50 years, likely because of better diagnostic modalities. Nonetheless, despite advances in imaging and surgical care, the current mortality rate ranges from 2%-20%.[1, 2, 7, 8, 11]  Not surprisingly, the mortality risk is greater in those with severe underlying comorbidities or uncontrolled sepsis. Differences in etiology (ie, iatrogenic vs noniatrogenic) do not affect the prognosis. The essential problem of spinal epidural abscess lies in the necessity of early diagnosis, as permanent neurological deficits and possible mortality can be avoided or reduced only with timely treatment.

Intracranial epidural abscess: With antibiotic and surgical management, intracranial epidural abscess carries a good prognosis, with an attributable mortality rate of less than 10%.

The neurological status of the patient at the time of diagnosis is the best predictor of neurological outcome, and morbidity is increased in both conditions when indicated surgery is delayed.[1, 2, 13] Comorbidities also often impact the outcome.

Sex

Most studies report that epidural abscess is more common in males than in females.

Age

Spinal epidural abscess can occur at any age. The median age of onset of spinal epidural abscess is approximately 50-60 years.

Intracranial epidural abscess is most common in the second and third decades of life.

The degree of neurologic recovery after surgery correlates with the duration and initial severity of the neurologic defect.

Spinal epidural abscess carries a mortality rate of 2%-20%; intracranial epidural abscess, about 10% (see Mortality). Eight percent of survivors were left paralyzed in one large 2017 series.[7]

Relapse after apparently successful therapy is unusual, but does occur in about 6.6% of cases.Recurrences tend to occur in those with a history of intravenous drug use,bowel dysfunction at presentation,or local spinal wound infection.[14]

Poor outcomes are more likely in patients with the following[1] :

• Multiple medical problems
• Prior spinal surgery
• Prior cervical or thoracic abscess location
• Thrombocytopenia
• Leukocytosis (>14,000 WBCs/µL)
• Persistently elevated inflammatory markers
• Infection with methicillin-resistant staphylococci
• Significant degree of thecal sac compression
• Sepsis

For patient education resources, see the Infections Center and Brain and Nervous System Center, as well as Brain Infection and Antibiotics.

Spinal epidural abscess

Most symptoms of a spinal epidural abscess are due to enlargement of the abscess and surrounding inflammation, which can lead to tissue compression and spinal cord ischemia. Onset of symptoms usually occurs within hours to days but may be more chronic in nature, presenting with weeks to months of symptoms. The microbiology often dictates the pace of progression.[1, 11]

Back or neck pain is the most common symptom in individuals with spinal epidural abscess, occurring in 70%-100% of cases.

The classic diagnostic triad of fever, spinal pain, and neurological deficits is present in only 8-15% of cases at first physician contact and must not be relied on for diagnosis. Significant back or neck pain coupled neurologic deficits, Staphylococcus aureus bacteremia, recent surgery/injections, or unexplained elevated C-reactive protein (CRP) or erythrocyte sedimentation rate (ESR) should prompt an MRI without delay to facilitate diagnosis.[8, 15]

If left untreated, the progression of symptoms is usually sequential and forms the basis for the staging of spinal epidural abscess (see Staging): (1) back pain; (2) radicular irritation; (3) motor weakness, sphincter dysfunction, sensory changes; and, finally, (4) paralysis. Note that this progression may occur very rapidly, and some symptoms may be skipped.

The patient's neurological status at the time of diagnosis is the most accurate predictor of outcome and prognosis.

Intracranial epidural abscess

The symptoms of intracranial epidural abscess are generally more acute but may be difficult to discern from the inciting process (eg, sinusitis, postoperative infection). When intracranial epidural abscess is combined with a subdural empyema, as is often the case, the course is compressed.[12, 13]

Signs and symptoms are due to both infection and the slowly expanding intracranial mass. Fever, headache, malaise, lethargy, nausea, and vomiting may be present. Intracranial epidural abscesses due to sinus infections can cause purulent drainage from the nose or ear.

Patients without a history of recent cranial manipulation who develop intracranial epidural abscess present with encephalopathy and focal neurological deficits. Most patients who have undergone craniotomy (67%) tend to be afebrile at presentation, and their neurological deficits are often less severe and less acute, with more than 90% showing evidence of wound infection.

Findings associated with spinal epidural abscess from multiple studies include the following[1] :

• Fever (range, 13%-95%; median, 32%) (However, note that many patients with spinal epidural abscess are afebrile.)
• Spinal tenderness (range, 17%-98%; median, 58%)
• Weakness of the extremities (range, 26%-87%; median, 40%)
• Sensory abnormalities (range, 13%-45%; median, 36%)
• Paralysis (range, 5%-39%; median, 27%)
• Reflex abnormalities (up to 40% of cases) (Early hyperreflexia may give way to diminished or absent reflexes.)
• Respiratory compromise (with cervical lesions)

Findings associated with intracranial epidural abscess include the following[12, 13] :

• Fever (However, fewer than half of patients are febrile, so this symptom is unreliable.)
• Headache (50%-73%)
• Altered mental status (44%-50%)
• Sinus tenderness (32%-90%)
• Focal neurological deficits
• Evidence of wound infection (>90% of patients who have undergone craniotomy)
• Seizure (4%-63%)

The microbiologic causes of spinal epidural abscess and intracranial epidural abscess are considered separately.

S aureus infection causes most cases of spinal epidural abscess. This is followed in frequency by streptococcal and Enterobacteriaceae infections. Coagulase-negative staphylococcal infections are observed almost exclusively in the context of recent spinal instrumentation or other medical procedures. The most common organisms that cause spinal epidural abscess include the following[1, 11, 16, 17] :

• S aureus (60%; up to 25% may be methicillin-resistant S aureus [MRSA]) ^[7]
• Enteric gram-negative bacilli, especially Escherichia coli (10%); their incidence is rising in some series
• Coagulase-negative staphylococci (3-5%), primarily involving spinal instrumentation or epidural anesthesia/injections
• Bacteroides species and other anaerobes (2%) ^[18]
• Pseudomonas species (2%)
• Streptococci, including Streptococcus viridans, group B streptococci, and pneumococcus (10%)
• Pasteurella multocida-few case reports ^[19]
• Mycobacteria, usually Mycobacterium tuberculosis (< 1% in Western countries but much more common in developing countries). Do not forget to send acid-fast bacilli (AFB) stains and cultures.
• Less-common organisms -Acinetobacter, enterococci, Actinomyces species, Nocardia species, ^[20]  Brucella species, and fungi, including Candida, Coccidioides, Aspergillus, ^[21]  Blastomyces, and Sporothrix species
• Exserohilum rostratum was a common cause in 2012 due to the use of contaminated methylprednisolone injections
• Polymicrobial (possibly 5%-10%)
• Unknown (6%-10%)

In intracranial epidural abscess, upper-respiratory bacterial pathogens predominate in sinus-associated disease, whereas nosocomial pathogens are of concern in cases that develop after craniotomy. The most common causative organisms include the following:

• Staphylococci, both coagulase-positive and coagulase-negative
• Streptococci, including anaerobic and microaerophilic species
• Aerobic gram-negative bacilli
• Propionibacterium acnes
• Other anaerobes
• Can be polymicrobial

Fever is helpful if present, but half or more patients are afebrile.

Respiratory compromise is possible if intracranial or cervical spinal epidural abscess is present.

Always check for spinal tenderness-top to bottom-pressing firmly on each spinous process.

Perform a full neurologic examination-mental status, motor, sensory, reflexes. The neuro exam needs to be repeated daily (or more often), as rapid changes in the examination may have major management implications.

Death or permanent neurologic sequelae occur in a substantial proportion of patients with epidural abscess, especially those who present with major neurologic deficits or sepsis. In a large 2017 series, 7% died and 8% were left paralyzed.[7]

Routine tests

The CBC count may reveal leukocytosis, left shift, thrombocytopenia, and anemia. Only about two thirds of patients who present with spinal epidural abscess have leukocytosis at the time of initial evaluation.

The ESR and CRP are almost invariably elevated; this is a nonspecific finding.

Always obtain blood cultures prior to antibiotics, as they are positive in 60% of cases.[1, 2, 7, 10]

Abscess fluid/operative material

Perform Gram staining and routine aerobic and anaerobic cultures on all aspirated or surgically obtained abscess fluid.

Special stains and cultures for mycobacteria and fungi are indicated.

PCR studies for bacteria, and in some cases fungi and/or mycobacteria, are often of great value. Treating epidural abscesses without identifying the offending pathogen(s) is extremely challenging, and PCR studies often save the day. 

Consider Brucella cultures and serologies when this is a possibility; alert laboratory personnel that Brucella may be involved so they can take precautions.

MRI is the cornerstone of diagnosis in both intracranial epidural abscess and spinal epidural abscess. MRI has the greatest diagnostic accuracy and is the method of first choice in the diagnostic process.[1, 22, 23] The sensitivity of MRI is 90%-95%, and its specificity also exceeds 90%. In some cases, MRI findings are indeterminate, necessitating a repeat of the study. Gadolinium enhancement increases sensitivity for detecting spinal epidural abscess, even in the absence of contiguous bony infection, and enables better differentiation between abscess and surrounding neurological structures. Always image the entire spine (cervical, thoracic, lumbar) when spinal epidural abscess is a consideration.

CT scanning with intravenous contrast may demonstrate fluid collections in the epidural space (see image below). CT scanning is the procedure of choice when MRI cannot be performed but may miss abscesses readily appreciated on MRI.

CT scan showing a lenticular-shaped intracranial epidural abscess.

When combined with myelography, CT scanning is a fairly sensitive tool to diagnose spinal epidural abscess, but it carries considerable risk, including introduction of infection, bleeding, nerve injury, and spinal shock. Myelography may underestimate the length of a spinal epidural abscess and carries a risk of paralysis.

Plain radiographs may demonstrate osteomyelitis or vertebral collapse. While these should be performed in all cases, they are never enough to establish the diagnosis.

Make every effort to establish a microbiological diagnosis. Blood cultures are positive in 60% of patients with spinal epidural abscess and are essential.[2, 7]

CT-guided needle aspiration may be used to obtain material for analysis.

Surgical specimens must be stained and cultured appropriately (see Causes). PCR studies are often helpful, especially in the culture negative epidural abscess situation.

Lumbar puncture generally is not indicated in spinal epidural abscess and carries the risk of spreading the bacteria into the subarachnoid space, with consequent meningitis. It is contraindicated in intracranial epidural abscess because of the high risk of cerebellar tonsillar herniation due to increased intracranial pressure. When obtained in spinal epidural abscess, lumbar puncture usually reveals a nonspecific parameningeal infection picture, with elevated protein levels, normal or slightly depressed glucose levels, and modest pleocytosis. Results may also be normal or indicative of frank bacterial meningitis. Culture results may be positive in up to 25% of cases, but almost all these patients have positive blood cultures.[2]

A staging system for the progression of spinal epidural abscess exists and may be of some diagnostic value, but it must be stressed that not all patients move sequentially through the stages, and that deterioration may be rapid.[1]

1. Back pain, tenderness, and fever
2. Radicular pain, reflex abnormalities
3. Sensory abnormalities, motor weakness, bowel and bladder problems
4. Paralysis, which rapidly becomes permanent without surgical intervention

Spinal epidural abscess

A combined medical-surgical approach, with emergent surgical decompression and drainage of purulent material, has been the standard approach to spinal epidural abscess. Antibiotic-based therapy, sometimes combined with CT-directed needle aspiration, has traditionally been used only in patients who are at prohibitively high risk of surgery or who have a fixed paralysis that lasts more than 48-72 hours and that is presumed to be irreversible.

In recent decades, wider use of antibiotic-based, nonsurgical therapy for spinal epidural abscess has been advocated,[24, 25, 26] condemned,[27, 28] and cautiously discussed.[29, 30, 31, 32, 33] The current literature on the subject consists largely of small case series and is inadequate to resolve the controversy.[1, 2, 34] Various prediction criteria advocated to predict failure of medical management[32, 33] have been published, but their value has been questioned.[35]

If medical therapy is to be used as initial therapy for spinal epidural abscess and surgery held in reserve, several caveats apply, as follows:

• The patient should have no neurological deficits. Patients with deficits experience better outcomes with prompt surgery. ^{[36, 37]}
• A culture-proven microbiological diagnosis should be available (from blood culture or aspiration). MRSA seems to be associated with higher failure rates than MSSA; be extremely cautious when contemplating medical therapy for MRSA epidural infections. ^[38]
• Stringent follow-up by both the primary team and neurosurgeons must be available and emergent surgery available, if needed.
• The physicians caring for the patient must be aware that rapid deterioration may occur at any time (the first 72 hours being most risky) and that even prompt rescue surgery may leave the patient with a neurological deficit that might have been avoided with surgery at first diagnosis. Failure rates with medical management in one recent study were 41%, with some these failures termed "catastrophic". ^[33] Other studies have found failure rates of 17%-38%. Cavalier follow-up of medically managed patients with epidural abscess can be disastrous; repeat MRI scanning and/or surgical re-evaluation must be available and performed on an emergent basis when the clinical picture changes.
• Risk factors for failure of medical therapy include any neurological deficit, sensory changes, diabetes, malignancy, bacteremia, age older than 65 years, and MRSA as the causative organism. ^{[32, 33, 39]} Patients with multiple risk factors for failure of medical therapy should have the traditional combined medical/surgical approach, with the surgery performed soon after diagnosis.
• A follow-up MRI is necessary within 2-4 weeks to evaluate for improvement with medical therapy.

Empiric antibiotic therapy should include coverage of gram-positive cocci, particularly staphylococci (including MRSA), and gram-negative bacilli. Vancomycin has been the standard agent for gram-positive infections, although linezolid,  daptomycin, or ceftaroline could be considered.[40] The third- and fourth-generation cephalosporins and meropenem offer excellent gram-positive (except MRSA) and gram-negative coverage in addition to good CNS penetration.

Pending culture and PCR results, a combination of agents (vancomycin plus cefepime or similar) is needed. Additional coverage may be needed if some of the less-common etiologic agents (see Causes) are suspected. Always tailor coverage once culture data are available; for example, nafcillin, oxacillin, and cefazolin are much better drugs for MSSA infections than vancomycin. Cefazolin is roughly as good as the anti-staphylococcal penicillins for MSSA spinal epidural abscesses. Note that failures due to the development of resistance have occurred when daptomycin alone was used for MRSA spinal epidural abscess.[41] Experience with ceftaroline for these infections is limited, but it may be a useful second-line MRSA agent.[42, 43]

Intracranial epidural abscess

A combined medical-surgical approach is used for intracranial epidural abscess. A craniotomy is usually performed. Empiric antibiotic therapy is similar to that described for spinal epidural abscess; since many of these infections result from prior interventions, the possibility of more-resistant nosocomial organisms must be considered. Vancomycin plus cefepime or meropenem would be good starting choices, with metronidazole added to the cefepime if anaerobes are a major concern.[40]

Prompt decompression is used to manage intracranial epidural abscess, as it is uniformly considered a neurosurgical emergency.

As discussed in detail above, most patients with spinal epidural abscess require urgent decompressive laminectomy; other surgical techniques may be preferred in certain situations.[44]  Biportal endoscopic spinal surgery has been used in selected cases as an alternative to standard open procedures.[45] In some patients without neurologic deficits, medical therapy might be cautiously attempted, recognizing that disastrous outcomes may ensue from this conservative approach (see Medical Therapy). CT-guided drainage might be helpful in some cases of posterior spinal epidural abscess, but the literature on this is scant.

Emergent consultation with a neurosurgeon is mandatory for surgical decompression and drainage of purulent material in patients with intracranial epidural abscess. Emergent surgical intervention is also needed in most patients with spinal epidural abscess, and prompt consultation and tight follow-up are mandatory in those in whom surgery is deferred (see Treatment). Consultation with an infectious disease specialist is strongly recommended for both diagnostic and therapeutic assistance.

Follow-up MRI should be obtained if any clinical deterioration is noted in patients with an intracranial epidural abscess or spinal epidural abscess. Follow-up MRI at 2-4 weeks should be performed in patients with spinal epidural abscess undergoing exclusively medical treatment to ensure the abscess has improved. It is unclear whether surgically treated patients with spinal epidural abscess who are doing well require follow-up MRI, as the MRI findings often correlate poorly with the clinical course.[46]

Follow-up with the neurosurgeon is always needed.

Follow-up with an infectious diseases specialist is necessary to monitor ongoing antibiotics.

Frequent neurologic examination is warranted during the postsurgical recovery period and is especially critical in patients undergoing medical treatment for spinal epidural abscess.

Fever, leukocytosis, or new neurologic deficit necessitates repeated imaging, and further (or initial) emergent surgical exploration may be required.[44]

Follow-up MRIs to evaluate spinal epidural abscess in patients who are doing well may not be helpful, as the findings may not correlate well with clinical course.[46]

Physical therapy may be necessary for individuals with a residual neurologic deficit.

In the United States, by law, any unstable patient must be stabilized to the extent possible, including consultation and surgery, if indicated, before transfer.

The duration of antibiotic therapy is not well defined, but 4-12 weeks is generally considered adequate. Concomitant osteomyelitis requires a 6- to 12-week course. A transition to highly bioavailable oral agents might be appropriate in some cases; rely on an infectious disease specialist consultant for guidance. (See Medical Treatment for empiric selections.) Monitoring the erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) value may be helpful, as control of infection is usually associated with normalization of the nonspecific markers. Note that some patients are cured despite stubbornly elevated CRP/ESR values, but stopping therapy with high or rising values should always give one pause.

Class Summary

Empiric antimicrobial therapy must be comprehensive and cover all likely pathogens. Antibiotic combinations, usually vancomycin or another MRSA agent plus a broad gram-negative agent, are recommended in both intracranial epidural abscess and spinal epidural abscess while awaiting culture data. This approach ensures coverage for a broad range of organisms and polymicrobial infections. Once organisms and sensitivities are known, antibiotic monotherapy is recommended.

Ceftriaxone (Rocephin)

Third-generation cephalosporin with fair gram-negative and gram-positive activity. Superior CNS penetration. Arrests bacterial growth by binding to one or more penicillin-binding proteins. Does not cover MRSA, Pseudomonas species, or resistant nosocomial enterics.Not a good choice for MSSA-cefazolin, oxacillin or nafcillin preferred.

Ceftazidime (Ceptaz, Fortaz)

Third-generation cephalosporin with broad-spectrum, gram-negative activity (including Pseudomonas species). Poor efficacy against gram-positive organisms and some resistant gram-negative organisms. Arrests bacterial growth by binding to one or more penicillin-binding proteins.

Meropenem (Merrem IV)

Bactericidal broad-spectrum carbapenem antibiotic that inhibits cell wall synthesis. Effective against most gram-positive and gram-negative bacteria, with excellent CNS penetration. Has slightly increased activity against gram-negative bacteria and slightly decreased activity against staphylococci and streptococci compared with imipenem, but much less likely than imipenem to cause seizures.

Metronidazole (Flagyl, Protostat)

Imidazole ring-based antibiotic active against various anaerobic bacteria and protozoa. Not active against any aerobes. Must be used in combination for most bacterial infections.

Vancomycin (Lyphocin, Vancocin)

Potent antibiotic directed against most gram-positive organisms and active against most Enterococcus species. Indicated in patients who cannot receive or have failed to respond to penicillins and cephalosporins or who have infections with MRSA or another susceptible gram-positive organism.

Nafcillin (Nafcil, Unipen)

A penicillin used almost exclusively for MSSA. Is not effective against MRSA infections. Do not use empirically when MRSA infection is possible.

Cefazolin (Ancef, Kefzol)

First generation cephalosporin with excellent MSSA coverage. No MRSA coverage and poor for many gram-negative organisms, so not used until culture results showing MSSA are in hand.

Cefepime (Maxipime)

Excellent gram-negative coverage with fair gram-positive coverage-not useful for MRSA.

Oxacillin (Bactocill)

Similar to nafcillin-used almost exclusively for MSSA. Do not use until cultures positive for MSSA and negative for MRSA as it has no MRSA activity.

Daptomycin (Cubicin, Dapzura RT)

Binds to bacterial membranes and causes depolarization of membrane potential, Active against most gram-positive organisms, but primarily used as as a MRSA drug instead of vancomycin. Resistance can rapidly develop in high innoculum infections, or when prosthetic materials are present.