• Background
• Clinical Diagnosis
• Investigation
• Treatment
• Non-Antimicrobial Treatment
• Empirical Antimicrobial Treatment
• Directed Antimicrobial Treatment (when microbiology results are known)
• Duration of Treatment
• Treatment Failure

A brain abscess is a focal, intra-cerebral infection, which usually begins as an area of cerebritis and develops into a collection of pus surrounded by a well-vascularised capsule. A brain abscess evolves through stages of early, late cerebritis followed by encapsulation. Intraventricular rupture of the brain abscess is associated with high mortality.

A subdural empyema is a collection of pus located inside the skull between the dura and arachnoid mater on the surface of the brain. It is usually unilateral and has a tendency to spread rapidly through the subdural space until limited by specific boundaries. Empyemas account for 20% of all intracranial abscess cases (Agrawal 2007). Empyemas also have an association with cerebral venous sinus thrombosis. At presentation the empyema is almost always the more life threatening pathology and needs to be managed first as an emergency to stabilise the patient but cerebral venous sinus thrombosis should be actively excluded once the patient is stable or as a differential in a deteriorating patient whose empyema has already been drained.

These infections are initiated when microorganisms are introduced into the brain tissue or subdural space following trauma; contiguous pericranial infection; meningitis or haematogenous dissemination from a distant infective focus. In the largest paediatric series (Tekkok 1992) infections involving middle ear, paranasal sinuses, penetrating trauma and congenital cyanotic heart disease were the most common predisposing factors. Haematogenous spread resulting from sepsis is less common.

There are no randomised controlled trials of therapy for children with brain abscess & empyema from which to draw recommendations. Available literature is largely retrospective, reviewing predisposing factors, microbiological features, treatment outcomes and prognostic indicators. Data on children is further limited as these series are based mainly on adults with a small number of children; subset data on children are not available in any of them. Hence the text that follows is based on current practice, expert recommendations and the limited literature.

Advances in neurosurgical techniques, newer antimicrobials and better imaging technologies have facilitated the diagnosis and management of intracranial pyogenic suppurations over past 20 years; however, they still remain potentially fatal central nervous system infections. Different factors had been described as influencing the outcome in studies including neurological status at admission, (intraventricular rupture of brain abscess) and multiple abscesses. Imaging severity based on number, location, extent of perilesional oedema and midline shift is also an important prognostic indicator (Demir MK 2007). Infancy is recognised as a risk factor for mortality (Tekkok 1992). Early recognition and management of predisposing conditions is important for improving the overall outcome.
Brain abscesses & empyemas are frequently polymicrobial, the most common etiologic organisms in clinical series have been microaerophilic Streptococci and anaerobic bacteria. Additional organisms such as Staphylococcus aureus and Enterobacteriaceae are also seen depending on the underlying source. Streptococci, Staphylococci and Proteus were the commonest organisms in one paediatric series (Tekkok 1992).

Back to top

CLINICAL PRESENTATION
The clinical features in patients with an intracranial abscess evolve with time and depend upon the host pathogen interactions and location of the brain abscess.
Headache was the most frequent symptom followed by fever and seizures. Classical triad of fever, headache and neurological deficit or absence of fever should not be used to exclude the diagnosis of brain abscess. Papilloedema is common, in 61% of Tekkok’s series. The presence of focal neurological findings (e.g., hemiparesis, hemisensory deficits, aphasia, and ataxia) depends on the location of the abscess, and these findings are seen in approximately one-third to one half of cases.
Younger children may present with non-specific features or evidence of central nervous system infection such as meningeal irritation or seizures, they may enter this treatment pathway via guideline for fever and seizure.

Subdural emyema may present in the same way as an abscess or meningitis at first but untreated patients go on to rapidly deteriorate and a suspected subdural empyema should be discussed urgently with the neurosurgical team. The commonest clinical presentation is a triad of fever, sinusitis and neurological deficits followed by a fulminant and rapid downhill course (Agrawal 2007).

Seizures
Seizures occur in 10–72% of patients with brain abscesses (and in up to 50–70% of patients treated with surgical excision) and up to 50% patients with subdural empyemas. In majority of children with brain abscesses, the onset of seizures is delayed, with only 50% occurring within the first year after treatment. Legg et al. reported that initial seizures appear to occur after a longer latency period in children younger than 10 years old. It has been reported that the mean latency period is ~ 3 years and anticonvulsant agents are generally efficacious in controlling seizures.

Back to top

Recommendation: CT head, ideally with contrast administration, is the imaging modality of choice.
[Evidence level B]

CT with and without contrast is the first investigation used in almost all cases, however it may not be possible on CT to:

• Detect a small subdural empyema
• Differentiate a subdural empyema from haematoma
• Differentiate a brain abscess from a tumour

In these cases depending on the clinical picture and degree of radiological doubt a MRI with contrast and DWI (ideally + MRV given association with cerebral venous sinus thrombosis) is usually required if the patient is stable enough,

All patients with suspected abscess or subdural empyema on CT should however be urgently discussed with the neurosurgical team once suspicion is raised before/whilst arranging MRI as they are at risk of rapid deterioration.

Recommendation: MRI head may be required in some cases, following discussion with radiology.
[Evidence level D]
CT scanning, preferably with contrast administration, provides a rapid means of detecting brain abscesses, the size, the number, and the location. MRI scanning is superior to CT and helps establishing the diagnosis by early recognition of cerebritis in a small number of cases where a CT scan may miss the diagnosis. Only 3/96 cases were detectable only on MRI (Hakan, 2006). An abscess appears as a ring-enhancing, space occupying lesion.
Brain abscess development can be divided into 4 stages: 1) Early cerebritis (1-4 days); 2) Late cerebritis (4-10 days); 3) Early capsule formation (11- 14 days); and 4) Late capsule formation (> 14 days). In the earlier phases, a CT scan performed without addition of contrast may show only low-attenuation abnormalities with mass effect. In later phases, a complete peripheral ring may be seen. On CT scans obtained after administration of contrast material, uniform ring enhancement is virtually always present in later phases. In early phases, the capsule will be difficult to visualize via conventional techniques and double contrast CT often is helpful in defining encapsulation of abscess. Metastatic tumours, high-grade gliomas, cerebral infarction, resolving cerebral contusion or haematoma, lymphoma, toxoplasmosis, demyelinating disease and radiation necrosis must be kept in mind as the differential diagnosis for brain abscesses, appearing as ring-enhancing lesions. 113 out of 130 cases (Tekkok 1992) were supratentorial and those related to congenital heart disease had a predilection for left parietal lobe. CT or MRI scan may also reveal an infected source such as a paranasal sinus or an ear infection.
Unfortunately, there are no laboratory data that are pathognomonic of brain abscess or empyema. WCC may be normal. CRP is elevated in up to 60% of patients.
Recommendation: The best way to make a microbiological diagnosis is by culturing of abscess material obtained at the time of surgery, which may be aspiration or excision.
[Evidence Level B]
Recommendation: Pus collected in a sterile universal container (NOT SWAB) should be sent to Microbiology for urgent microscopy, culture and sensitivity.
[Evidence Level C]
Recommendation: Blood cultures should be obtained when a brain abscess is suspected (prior to commencing antimicrobial therapy).
[Evidence Level B]
Haematogenous spread may be the mechanism of CNS infection and a positive blood culture result may help guide therapy and subsequent investigations.
Recommendation: Lumbar puncture (LP) is contraindicated prior to intracranial imaging in any child in whom a brain abscess/empyema is suspected because of the potential for CNS herniation.
[Evidence Level C]
Role of imaging in monitoring response to therapy
Recommendation: Provided the child remains stable, repeating CT or MRI scans at weekly intervals for 2 weeks and then at fortnightly intervals for a further 1-month to enable re-accumulation of the abscess to be detected at a pre-clinical phase.
[Evidence level D]
Scans alone are unreliable in measuring response to treatment since radiological changes lag behind both the reduction in size of the cavity and the clinical response. Complete resolution of the abscess and associated abnormal contrast enhancement may take up to 12–16 weeks, and a small area of residual contrast enhancement may be present for up to 6 months after antibiotic therapy alone or in combination with stereotactic aspiration or surgical drainage via craniotomy. The size of the abscess decreases in 1–4 weeks with antibiotic therapy alone or in combination with surgical aspiration, and 95% of abscesses that resolve with antibiotic treatment alone demonstrate a reduction in size within a month. (Brook, 2004)

* Abbreviations: CE = contrast enhancement

Back to top

Recommendation: The treatment of brain abscesses should be a team approach, with collaboration between a Microbiologist, Neurologist, Neuroradiologist and Neurosurgeon.
[Evidence level C]

Souce control
If the source of the abscess/empyema is an active infection then whilst neurosurgical drainage normally takes priority as the abscess or empyema is acutely life threatening due to the risk of cerebral herniation, aggressive source control is also required. In ENT or maxillofacial infections this usually means a washout by the respective speciality under the same anaesthetic.

Surgical management
Operative management provides therapeutic and diagnostic benefits in patients with brain abscesses and subdural empyemas. There is relief of mass effect for larger empyemas and encapsulated abscesses, and cultures can be obtained. Typically, abscesses > 2.5 cm require surgical intervention. The optimal management of intracerebral abscesses varies according to lesion size, location, and stage. Early aspiration is advocated in infants because of the propensity for early seizures with meningitis and hydrocephalus, all of which may indicate a poor prognosis. Perioperative and postoperative antiepileptic therapy is generally recommended, and the duration of this therapy should be tailored individually taking into account pharmacological interactions with antimicrobials. (Sheehan 2008).

Intracranial empyema drainage
Empyemas that are extradural (outside the dura mater) and small can sometimes be drained via primary source control (eg sinus drainage or mastoidectomy) but large or subdural empyemas are more likely to need neurosurgical evacuation via burrhole or craniotomy.

Abscess drainage
There are various options for draining intracranial abscesses including burrhole aspiration (which may be anatomically planned, image guided or sterotactic); craniotomy and open abscess drainage and craniotomy and abscess excision. The best surgical approach will be decided by the neurosurgical team on a case by case basis.

Intraventricular rupture of abscess
Intraventricular rupture of an abscess is associated with a high mortality rate and will likely correlate to a clinical deterioration in the patient due to hydrocephalus and ventriculitis. In such cases in addition to draining the primary abscess one or more external ventricular drains may be required to treat hydrocephalus and allow the administration of intrathecal antibiotics in addition to intravenous antibiotics.

Management of multiple brain abscess
Stereotactic aspiration or surgical excision should be performed for abscesses larger than 2.5 cm or smaller lesions that are causing a significant mass effect. If there are no abscesses larger than 2.5 cm in diameter, then the largest abscess should be aspirated for the purpose of cultures. The duration of the antibiotics course is usually 6–8 weeks and longer for immunocompromised patients. Serial imaging is conducted on a weekly basis or immediately if there is a decline in the patient's neurological status. A decline in the patient's clinical status, enlargement of the abscess after 2 weeks of antibiotics, or lack of reduction in the size of the abscess after 4 weeks of antibiotic therapy should prompt consideration of further surgery. (Mathisen 1997)

Back to top

Recommendation: Initial antimicrobial choice is empirical and should be tailored to cover the most likely pathogens in individual cases depending upon the location of the abscess/empyema and predisposing focus (dental, paranasal sinuses, otogenic etc.), according to Table 2.
[Evidence level D]
Recommendation: Wherever possible (e.g. in stable children in whom early surgery/drainage is planned) empirical antimicrobial therapy should be started after surgical drainage, or, if this is not possible, after blood cultures have been taken.
[Evidence level D]
Recommendation: Broad spectrum empirical antimicrobial regimens should be tailored to organism isolated from pus or blood cultures i.e. converted to “directed antimicrobial therapy” whenever possible (see Table 3).
[Evidence level D]
Antimicrobial therapy
No randomized controlled trials of therapies for brain abscess have been conducted because of the relative infrequency of this infection. Appropriate antimicrobial therapy for brain abscess depends on choosing antibiotics that are able to penetrate the abscess cavity and have activity against the suspected pathogens. Although there are considerable data on CSF antimicrobial levels in cases of bacterial meningitis, there is still limited information available on antimicrobial penetration into brain abscess cavities. The blood-CSF and the blood-brain barriers are histologically different and a high degree of CSF penetration does not guarantee penetration into the brain tissue.

* Use Ceftazidime as third generation cephalosporin if Pseudomonas infection suspected.
Contact Medical Microbiologist for advice if any risk factors for resistant gram negative infections (e.g. previous colonisation, intensive care unit stay, transfer from outside UK)
** Please discuss antibiotic options with Microbiology in cases of penicillin allergy.
There have been a number of more recent case reports demonstrating successful non-operative treatment of brain abscess with antibiotics alone. This approach may be appropriate for clinically stable patients who are poor candidates for surgery or for patients with surgically inaccessible lesions. Small lesions (2 cm) located in the better-vascularised cortical areas are more likely to respond to antibiotics alone. Medical treatment alone should not be used when the diagnosis is in doubt or when pathological confirmations are not available. Serial CT or MRI scans are crucial because abscesses may enlarge despite antibiotic treatment. If neurological deterioration occurs as a consequence of mass effect, surgical removal may become necessary.
Once a patient has stabilised and is improving oral antibiotics can be considered in a patient when: Evidence Level C

• CRP shows downward trend if the patient is otherwise well
• The patient has no signs of fever
• Appropriate oral agents are available
• The patient can tolerate oral antibiotics

This would normally only be considered once imaging is showing an improving trend.
Please contact Microbiology to discuss oral antimicrobial options.
Intrathecal administration of antibiotics is not routinely recommended (apart from intraventricular rupture of brain abscess)

Back to top

Recommendation: Broad spectrum empirical antimicrobial regimens should be tailored to organism isolated from pus or blood cultures i.e. converted to “directed antimicrobial therapy” whenever possible (see Table 3).
[Evidence level D]

**Dose adjusted according to estimated creatinine clearance; need to monitor Vancomycin trough levels.
Treatment of Primary Focus
To minimise the risk of recurrent or non-responsive intracranial infection, any identifiable primary source requires aggressive treatment. This may include surgery for paranasal sinuses, middle ear or dental sepsis, physiotherapy and echocardiography in patients with a cardiac source. The timing of such interventions does not need to coincide with intracranial surgery but should be undertaken in an expert, timely fashion.

Back to top

Recommendation: Duration of therapy should be determined by clinical progress, CT findings and whether surgical drainage has been undertaken.
[Evidence level C]
Recommendation: A minimum of 4 weeks treatment (usually 6-8 weeks) is required if the abscess is treated with antibiotics alone (demonstrate resolution of ring enhancing lesions on CT before stopping antibiotics).
[Evidence level C]
Recommendation: Give 3-4 weeks if abscess has been excised and 4-6 weeks if abscess has been aspirated or excised, provided the clinical response is good.
[Evidence level C]
The appropriate duration of antimicrobial therapy for brain abscess remains unclear. A 6–8 week course of parenteral antibiotics has traditionally been recommended provided the aetiological organisms are susceptible and that adequate surgical drainage can be established. More recently, shorter durations of antibiotic therapy have been proposed based on correlation between clinical progress and CT findings.

Back to top

Recommendation: Discussion between neurosurgery and microbiology is recommended.
[Evidence level D]

Back to top