Clinical Context

A recent review of large inpatient databases found that the rate of admission for all identifiable intracranial focal suppurative infections secondary to sinusitis or otitis (including epidural abscesses, subdural empyemas, and brain abscesses) was 2.74 to 4.38 per million children in the United States.1 A busy neurosurgical service in the developed world with two pediatric neurosurgeons can therefore expect to see two to four of these patients in a given year.1,2 Several key events have shaped our current approach to these lesions. The introduction of antibiotics to patient care reduced reported mortality from between 60 and 80% to between 20 and 40%.3 This rate has been further reduced to between 0 and 10% in modern series both by the availability of computed tomography (CT) and magnetic resonance (MR) imaging and by improvements in antibiotic therapy.1,2,4,5

Extra-axial abscesses have a bimodal distribution based on patient age. This finding is consistent throughout the literature and was recently reiterated in a review of 70 patients treated for extra-axial abscesses at The Hospital for Sick Children in Toronto, Canada, between 1995 and 2009.2 Fifty percent of these children were older than 11 years of age, and 21% were below 1 year of age. In children younger than 5 years of age, the extra-axial abscess was most often the consequence of meningitis or otitis media. All patients with post-meningitis extra-axial abscesses were younger than 1 year of age. All patients with post-sinusitis abscesses were older than 7 years of age because the frontal sinus is not pneumatized in younger children. Central nervous system (CNS) infection complicates 3 to 4% of hospital-admitted cases of sinusitis.6 Valveless mucosal veins penetrate the inner lamina of this sinus and allow communication with both the diploë and the dura mater.7 Unlike other etiologies, postoperative extra-axial abscesses were seen to span these age groups.

Infants who develop a post-meningitic abscess (often subdural) have signs and symptoms of meningitis that include a bulging fontanelle, and their condition often fails to improve despite adequate antibiotic therapy for their meningitis.8 In the older child, Pott puffy tumor is a common finding in patients who have developed an intracranial extra-axial abscess; in the Toronto series, this was found in 16 of 38 (42.1%) patients with post-sinusitis abscesses.2,9 In the absence of a Pott puffy tumor, the clinical presentation does not follow a consistent pattern; although headache, fever, emesis, and meningismus are often present, the specificity of these findings is far too low to be used for effective decision making.8 However, the clinical presentation, in conjunction with a high degree of suspicion, often prompts imaging studies by the admitting and subsequently the consult-seeking service, whereupon intracranial mass lesions are revealed if present. Routine blood work is rarely useful in diagnosing these infectious lesions. Neither a leukocytosis nor a left shift is always present. The erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) level can be useful for monitoring response to therapy for established lesions but lack the positive or negative predictive value needed to be helpful in diagnosis.8 Blood cultures are similarly of low yield, though, because of their ease of collection, they should be obtained on the slight chance of assisting in the microbial identification. CSF analysis for these lesions is neither sensitive nor specific, and results will often be normal.8

The identification of most extra-axial abscesses on CT and MR imaging is straightforward. On CT scans, the extra-axial abscess will show a hypodense collection of pus with enhancement along both the cortical and the dural border. T1-weighted MR imaging will show the pus collection to be iso- to hypointense, and on T2-weighted imaging it will be hyper- to isointense.10 On post-contrast MR imaging, there is usually enhancement of the outer border on the medial, lateral, or both margins. Infants can pose a radiographic challenge in differentiating a post-meningitic subdural empyema from a sterile reactive subdural effusion. Although reactive subdural effusions can occur in 33% of infantile patients with meningitis and usually do not require intervention, subdural empyemas occur in only 1%.11 On ultrasound, the reactive subdural effusion is usually anechoic, with the thickened hyperechoic inner membranes and echogenic cerebral sulci characteristic of meningitis. Chen et al were able to use ultrasound to correctly distinguish subdural empyema from reactive subdural effusion in 15 of 16 lesions. They found that subdural empyema tends to be anechoic with prominent traversing fibrinous strands in the early stages and hyperechoic in its entirety with frequent loculi as it matures.12 CT scans of a subdural empyema can vary widely and lack high specificity or sensitivity, although some authors in the developing world have used CT as the sole radiographic modality because of its accessibility and low cost.10,13,14 MR imaging is probably the best modality to differentiate these entities. Wong et al reported on 10 patients with a total of 12 lesions. Of the 12 lesions, 10 were subdural empyemas and 2 were reactive subdural effusions; MR images of 9 of the patients with subdural empyemas revealed restricted diffusion, while one image demonstrated mixed signal on diffusion-weighted imaging. Both patients with reactive subdural effusions demonstrated low signal intensity on diffusion-weighted imaging.15 In older children, imaging is more straightforward, with CT the nearly universal modality used to evaluate children with a suspected extra-axial suppurative process. MR imaging is often used to clarify anatomic detail or to characterize the abscess better in cases in which it will be followed with conservative treatment. In the previously cited Toronto series, MR imaging was used in only 12 of 38 patients with sinusitis-related extra-axial abscess; this may be related to the fact that the vast majority of these cases (64 of 70 patients) were managed operatively, necessitating only CT localization rather than the subtle definition of improvement with medical therapy.2

Abscesses of the brain parenchyma tend to have two main causes. Hematogenous spread occurs in children with congenital cyanotic heart disease and a left-to-right shunt, which bypasses the reticuloendothelium of the lung, or an infectious focus elsewhere in the body. Contiguous spread can occur, as with extra-axial abscesses, from sinus or middle ear disease. Cases arising from trauma or as a postsurgical complication are less common, while abscess in the context of immunosuppression is an increasingly frequent clinical situation.8 Some 10 to 37% of intraparenchymal abscesses are idiopathic.16 The clinical presentation of an intraparenchymal abscess lacks both sensitivity and positive predictive value for the disease. The classic triad of headache, fever, and neurologic deficit is present in only 17 to 30% of cases.8,17 As with extra-axial abscesses, the clinical blood count (CBC), ESR, CRP level, and CSF evaluations are seldom helpful beyond tracking the effect of therapy in patients with a diagnosis. An exception to this rule is in the patient in whom the abscess has ruptured into the ventricle, who will usually have a positive CSF evaluation. Blood cultures return relevant results approximately 30% of the time and should be obtained if there is suspicion of this diagnosis.18

CT or MR imaging should be obtained with contrast and reveal either an edematous region with mass effect in the cerebritis stage or a ring-enhancing lesion with a hypodense interior in the later stages. Early abscesses can have a radiographic appearance similar to that of an ischemic stroke on CT without contrast; therefore, the index of clinical suspicion should guide the need for repeat CT with contrast or enhanced MR imaging.19 Characteristics that distinguish mature abscess from tumor are gas within the center of the lesion, a rim of less than 5 mm (which tends to be thinner than that of a brain tumor), and ependymal enhancement that can be associated with ventriculitis or ventricular rupture. Diffusion-weighted images show restriction (hyperintensity) in the interior of the abscess, with a dark absolute diffusion coefficient (ADC).18 If needed, further differentiation from a tumor can be made with perfusion MR imaging, which demonstrates hypovascularity in an abscess capsule and hypervascularity in a tumor capsule.8 A radioactively labeled leukocyte scan can also be used to diagnose an infectious process in uncertain situations.20 In infants with an open fontanelle, ultra-sound can be used to diagnose and follow intraparenchymal abscesses.21

Medical Management

Although clear parameters for medical management of intracranial abscesses (epidural, subdural, and intraparenchymal) are still not defined, the method of treatment has been under investigation since 1970.4 The widespread availability of both potent antibiotics and improved intracranial imaging has expanded the role of medical management. In most cases, intracranial abscesses that are less than 2.5 cm can be managed medically if they present in otherwise healthy patients without focal deficits, altered sensorium, or the imminent threat of either. In sick, high-risk patients or in patients with deep-seated abscesses, neurologic deficits may be unavoidable and can worsen with surgery.

Few specifics exist for the medical management of epidural abscesses. Nathoo et al first reported on this strategy in 1999, managing six patients without fever with antibiotic therapy and serial imaging. No information is available on lesion size or whether there was associated sinus disease that was treated by the otolaryngologic service.22,23 Heran et al contributed to the literature on this topic in 2003, treating four patients who had isolated epidural abscesses with sinus drainage and antibiotics without intracranial intervention. There was no evidence of bony erosion, which would have allowed spontaneous drainage of the intracranial component into the extracranial space. None of these patients had focal neurologic deficits or an altered sensorium, and the average size of the abscesses was 2.9 × 2.6 × 1.4 cm, with a maximum size 3.5 × 3.5 × 1.2 cm in one that was infratentorial.24 At our institution, we routinely treat epidural abscesses as a medical disease to great effect and rarely operate unless there is a decreased level of consciousness or focal neurologic deficits attributable to the lesion. Decisions regarding the treatment of primary sinus disease are left to the discretion of the otolaryngologists.

The medical management of subdural empyema was first explored in an article from 1979 by Rossaza et al, in which a child with an interhemispheric empyema and foot weakness was cured clinically and radiographically with antibiotics alone.25 Steroids were used in this case to ameliorate the foot weakness. Leys et al provided more support for medical management in the 1980s, finding that individuals as young as 2 years of age with empyemas up to 44 mm thick in the supratentorial space, with an average Glasgow Coma Scale score of 11 (range, 7 to 13) and even with a severe neurologic deficit, could be medically treated with a good clinical outcome. Although the length of stay was shorter and the infectious organism was more often identified in both the aspiration and craniotomy groups of patients, medically treated patients had the same mortality rates and a lower probability of persistent neurologic deficits and seizures than those treated surgically.26,27 Similar to Heran et al, Leys found that minor deterioration in neurologic status within 24 to 48 hours of the start of medical treatment often resolved without the need for surgical intervention. A few modern series have made reference to medically managed empyemas, but they do not include the details of these patients.2,28 Salunke et al did describe one patient treated successfully with antibiotics alone who had a thin subtentorial empyema.28 At our institution, we have even had success treating subdural empyemas of the posterior fossa with antibiotic therapy. Recently, we successfully treated an 8-day-old infant with Currarino triad; she had multiple infratentorial subdural empyemas that measured 1 and 0.6 cm, and at no time did she demonstrate a neurologic deficit. After 6 weeks of intravenous antibiotics, all empyemas had resolved, and she remained without neurologic deficit ( Fig. 17.1 ).

Decisions regarding the medical management of intraparenchymal abscesses are similar to those for other intracranial abscesses and include abscess size, neurologic status of the patient, general medical condition of the patient, and the need for organism identification. In addition to these factors, the location of the intraparenchymal abscess must be considered because the indications for operating within eloquent brain must be justified. No specific criteria for the appropriate size of an intraparenchymal abscess have been established that will result in its successful medical treatment. A comprehensive review of the literature has provided a thoughtful, well-supported approach to patients with intraparenchymal abscess.29 Patients with intraparenchymal abscesses smaller than 2.5 cm in maximum diameter are the best candidates for medical management, especially where their overall neurologic condition is good and microbiological speciation has been determined from another source. Although well supported in the literature, the size of 2.5 cm or less may be an unduly conservative requirement for medical management. In a review of reports of the medical treatment of intraparenchymal abscesses, Bamberger found that 100% of 113 brain abscesses less than 5 cm in maximum diameter and 74% of 30 intraparenchymal abscesses 5 cm or more in diameter were successfully treated medically.30 At our institution, we routinely treat intraparenchymal abscesses more than 2.5 cm in maximum diameter medically, with only a rare conversion to operative management. Multiple abscesses can also be treated medically, with surgical excision of intraparenchymal abscesses larger than 2.5 cm or of smaller ones causing symptomatic mass effect. When a patient is being treated medically, surgery should be reconsidered if the patient′s condition deteriorates clinically or if clinical or radiographic improvement is not observed within 1 to 2 weeks.

Surgical Management

Although the indications for medical management of intracranial focal infections have expanded, the large majority of these lesions are handled operatively. In their large review of 70 patients with extra-axial abscesses, Gupta et al operated on all but six patients.2 Similarly, most modern series reporting on extra-axial abscesses indicate that surgery is overwhelmingly the first choice of treatment.14,28,31,32 The main cause for concern in approaching an extra-axial abscess operatively is the extent of exposure. Although it is clear that extra-axial abscesses can be treated with either burr hole drainage or limited craniotomy, the rate of recurrence—requiring reoperation to clear the infection—is potentially higher after burr hole drainage in larger extra-axial abscesses and in those with evidence of loculi.5,16,22,31,33–37 Parafalcine or paratentorial empyemas may pose a higher risk for recurrence after burr hole drainage than after craniotomy.28 In contrast, a series of post-meningitis empyemas demonstrated that these lesions might preferentially be treated safely with burr hole drainage rather than with craniotomy.31 Certainly, in a decision to use a surgical approach to an infection, the clinical context must be considered. Those patients with significant focal or global neurologic deficits should be considered candidates for craniotomy, with craniectomy more appropriate for those at risk for postoperative hemispheric swelling. For the significant number of patients who have an associated sinusitis or mastoiditis, an otorhinologic consult should be obtained. If possible, definitive intracranial management should occur at the same time as otorhinologic treatment because several authors have reported lower rates of intracranial recurrence with concurrent management.2,7,22

Issues regarding management of the bone flap arise in these cases, with most authors advocating immediate replacement of the bone flap unless frank, extensive osteomyelitis is present. A report addressed the issue of devascularized bone flaps in the context of extra-axial abscesses without osteomyelitis.38 In this series of 14 patients, all had successful immediate reimplantation of the bone flap following craniotomy for an extra-axial abscess. One of these patients had tetralogy of Fallot, and two had undergone chemotherapy and radiation therapy following resection of brain tumors. In each case, the surgeons removed all soft tissue from the bone flap, scrubbed the flap for 3 to 5 minutes in iodophor or bacitracin solution, then soaked the flap in an iodine or bacitracin solution until reimplantation. In addition, the surgical bed was scrubbed abrasively with surgical sponges, and the bone edges were débrided, with care taken not to injure the exposed brain. Two other reports used an implanted irrigation–drainage system to salvage devascularized bone flaps in the context of postsurgical infection but did not comment on the degree of osteomyelitis present in the flaps.39,40 Unfortunately, there are no clear data on the possibility of immediate reimplantation of a bone flap affected with frank osteomyelitis, and the risks and benefits of aggressive débridement and immediate replacement of the osteomyelitis flap must be left to the surgeon. Of note, in the report by Gupta et al on 12 patients receiving a craniectomy for osteomyelitis, six did not require a cranioplasty because in situ bone growth with a good cosmetic result was present. No ages were provided for these patients.2

Although abundant literature is available on the management of intraparenchymal abscesses, it lacks patient stratification, clear management protocols, and consistent outcome measures, decreasing the usefulness of this body of information. In a review from Italy, the authors made several useful conclusions regarding intraparenchymal abscesses.29 First, the choice of a surgical approach that includes either a burr hole or a craniotomy did not appear to impact outcome. Instead, the patient′s initial neurologic condition and the rapidity with which therapeutic intervention is initiated are far more important prognostic factors. Those determinants of operation type are influenced by the preference and abilities of the surgeon and whether the patient will be able to tolerate the proposed procedure. It might be more reasonable to aspirate those intraparenchymal abscesses that are located deep within eloquent areas of the brain or that are small or multiple, even if they are recurrent. Those intraparenchymal abscesses that are superficial, are located in the posterior fossa, or result from posttraumatic or postoperative complications might better be treated by a craniotomy than by burr holes. It is not clear whether the size of the abscess and the neurologic status of the patient are independent determinants of what type of management is required. We recently encountered a 9-month-old girl with a 3-week history of progressive lethargy and decreased appetite. On her evaluation in the emergency department, she was fussy but consolable and had a bulging fontanelle. MR imaging revealed multiple abscesses occupying nearly the entire left hemisphere. Each abscess was aspirated through a separate burr hole with a ventricular catheter and a syringe that had a Luer connector. A catheter was left overnight in the largest abscess, and that abscess was reaspirated the next day. The patient was then treated with antibiotics and had a good clinical outcome ( Fig. 17.2 ).