Reddy, Pavani MD*; Anderson, Evan MD*; Naidich, Michelle J. MD†; Miller, Frank H. MD†; Noskin, Gary A. MD*
A 43-year-old man with a history of squamous cell carcinoma of the head and neck was admitted to the hospital with 24 hours of fever and confusion. The patient, a former smoker, originally presented with a left-sided neck mass 8 years ago. A modified radical neck dissection was performed, revealing poorly differentiated squamous cell carcinoma. The patient refused adjunctive radiation therapy at the time of diagnosis and resection. The tumor recurred one year later, at which time the patient underwent a second surgical resection followed by postoperative radiation. Five years ago, the patient developed a recurrence in his left cheek. He subsequently underwent left total parotidectomy, wide resection of the buccal lesion, partial maxillectomy, marginal mandibulectomy, dental extraction, and split-thickness radial skin graft reconstruction. Despite these aggressive surgical interventions, the patient developed recurrent maxillary sinus squamous cell carcinoma, not amenable to further operative measures. He refused further radiation and chemotherapy.
Upon admission, the patient's wife reported several days of thick yellow drainage from the surgical resection site. In the 24 hours before presentation, she noted that he was intermittently febrile, increasingly confused, and less responsive. Upon physical examination, the patient's temperature was 96.9°F, and pulse rate was 73 beats per minute; respiratory rate was 20 breaths per minute, and blood pressure was 101/63 mm Hg. He seemed lethargic but was arousable and responsive to simple commands. Consistent with prior examinations, the patient had left eye proptosis and a 6 × 8-cm left buccal defect with clean margins that communicated with the oropharynx. The alveolar ridge was absent, and so were the portions of the hard and soft palate. Portions of the mandibular angle and ramus were exposed, but the tongue and the floor of the mouth were intact. Of note, the open wound was malodorous, and a moderate amount of thick yellow pus was noted at the mandibular angle. In addition to preexisting cranial nerve deficits from prior surgeries, neurological examination demonstrated a new receptive aphasia.
Laboratory results revealed a serum white blood cell count of 12,700 cells/μL, with 94% polymorphonuclear cells, hemoglobin of 9.6 gm/dL, and platelet count of 315,000 cells/μL. Chemistry values were notable for a serum sodium 128 mEq/L. Urinalysis was unremarkable. A computed tomography (CT) scan without contrast was performed (Fig. 1). The following day, gadolinium-enhanced magnetic resonance imaging (MRI) demonstrated a temporal lobe lesion with associated edema (Figs. 2, 3). Based on these clinical and radiographic findings, the patient was placed empirically on intravenous vancomycin and cefotaxime.
On hospital day 3, the patient underwent stereotactic aspiration of the lesion. Three milliliters of purulent material was obtained. Gram stain revealed many gram-negative bacilli. Cultures subsequently grew Peptostreptococcus, Fusobacterium, and Bacteroides species, in addition to rare Staphylococcus epidermidis. The patient was discharged with intravenous ceftriaxone and oral metronidazole for 8 weeks.
BACKGROUND AND PATHOPHYSIOLOGY
Brain abscess is a focal suppurative process within the brain parenchyma, progressing from a localized area of cerebritis to a well-vascularized, encapsulated collection of pus.1 In developed countries, brain abscesses account for 1% to 2% of all intracranial space-occupying lesions.2 This review will focus on the clinical features, radiographic presentation, and management of brain abscesses, with special attention to infections of otolaryngologic origin resulting in contiguous spread.
In 70% to 80% of cases, a risk factor for brain abscess formation can be identified. Most cases (19%-65%) are caused by contiguous spread from dental infections, sinusitis, otitis media, or mastoiditis.3 Valveless emissary veins filtering the paranasal sinuses, middle ear, and teeth allow for retrograde transmission of bacteria into the venous drainage system of the brain. Contiguous spread may also occur via the internal auditory canal, cochlear or vestibular aqueducts, sinuses, mastoid cavity, or temporal bone suture lines. As a consequence of local sinusitis or otitis, most brain abscesses are localized to the frontal (12%-43%) or temporal (14%-42%) lobes.4-6 On rare occasion, cerebellar abscesses may also occur as a consequence of otogenic infection.3,6 One large series suggests that contiguous spread, as an etiology for brain abscesses, has declined from 55% to 17% since the mid-1980s.7 Reasons for this decline in incidence are unclear, however, a more aggressive approach to the treatment and diagnosis of otolaryngologic infections has been implicated.3
More than half of otogenic brain abscesses are polymicrobial in origin, consisting of microaerophilic and anaerobic streptococci, Haemophilus, Bacteroides, Fusobacterium, and Peptostreptococcus species.1,6,8,9 Although the isolation of anaerobic flora from brain abscesses varies, recent studies suggest that 33% to 41% of monomicrobial infections are caused by anaerobes. Increased identification of these organisms may be attributed to improved anaerobic transport techniques and prompt plating of infected material on appropriate media.6,10,11 Brain abscesses arising from an otic source may also include members of the family Enterobacteriaceae or Pseudomonas aeruginosa.
Most of the remaining cases of brain abscesses are attributed to hematogenous spread (13%-45%) and antecedent head trauma or neurosurgical procedures (10%-26%).1,6,8,12 Brain abscesses caused by hematogenous spread are often the consequence of congenital heart disease or pulmonary infection; however, primary urinary tract and intra-abdominal sources have been reported.9,13 One large series identified bacterial endocarditis, most often caused by Streptococcus viridans or Staphylococcus aureus, as the underlying source in 7% of adult cases.4 Pediatric cases represent 25% of all brain abscesses and are most commonly caused by hematogenous spread.14
Staphylococcus aureus, isolated in 10% to 31% of brain abscesses, is a common agent of infection caused by trauma or penetrating brain injury.15,16 Significant wound contamination in such cases has also led to infection with facultative gram-negative bacteria and Clostridium species.17 An estimated 5% to 20% of brain abscesses are postoperative in etiology.13 Such postneurosurgical infections are most often caused by S. aureus or S. epidermidis.15,16
Despite improvements in culturing techniques, abscess cultures may remain sterile in 9% to 63% of the cases.3,11 Although a presumptive source of infection may be determined in some cases, up to 20% to 41% of brain abscesses remain cryptogenic, with one study suggesting that the incidence of such cases is increasing.3,5,13
The duration of symptoms in patients with brain abscesses varies; patients may experience symptoms for as little as several hours or as long as 3 months before admission, with a median duration of 7 to 14 days.6,12,18 Less than 50% of the patients with bacterial brain abscess present with the classic triad of fever, headache, and neurological deficit.4,15 Of the infected patients, 44% to 59% present with headaches, 39% to 64% with fevers, 16% to 46% with seizures, and 36% to 54% with behavioral changes or focal neurological deficits.1,4,5,8,12 Associated meningismus and papilledema are less common.7,16
Focal symptoms on presentation correspond to affected regions of the brain. In the largest known series of brain abscesses to date, patients with temporal lobe involvement presented with speech disturbances and hemianopsia in 20% and 31% of the cases, respectively. Patients with cerebellar abscesses demonstrated ataxia or nystagmus in more than half of the cases. Notably, more than one third of the 400 study patients presented without localizing signs.5
Upon presentation, most patients have an elevated serum white blood cell count; however, other laboratory results are usually unrevealing.4,6 When performed, erythrocyte sedimentation rate was elevated in 33% to 63% of the patients.5,13 Because of frequent clinical and radiographic evidence of increased intracranial pressure, lumbar puncture is rarely performed. Yang5 and Xiao et al4 reported lumbar puncture findings for 173 patients with brain abscess: 117 (68%) patients had elevated intracranial pressure, and 136 (79%) had a pleocytosis (>10 cells/μL). Cerebrospinal fluid culture is rarely positive for an etiologic agent; however, in rare cases, the isolation of an organism was associated with communication between the abscess and the ventricular system.6,12,18
Marked improvement in the mortality caused by brain abscesses has been associated with the advent of CT imaging. One center noted that the mortality rate fell from 44% to 0% after CT imaging became available.19
The radiographic appearance of brain abscesses changes as the lesion progresses from initial infection to a mature abscess. In 1981, Britt et al20 described 4 stages of histologic and CT abnormalities seen in an animal model of brain abscess after intracranial injection of α-hemolytic streptococcus. In early cerebritis (days 1-3), acute inflammatory cells are noted between the developing necrotic center and the surrounding brain tissue, with marked cerebral edema surrounding the lesion. Contrast-enhanced CT scan demonstrates partial ring enhancement. Late cerebritis (days 4-9) is characterized by the appearance of fibroblasts and extensive cerebral edema. Contrast-enhanced CT scan reveals a well-formed ring that diffuses into the center of the capsule on delayed imaging. In the early capsule stage (days 10-13), a collagen capsule starts to form, whereas the size of the necrotic center and surrounding edema decreases. Contrast-enhanced CT scan demonstrates ring enhancement with less contrast diffusion. Finally, in the late capsule stage (day 14 and beyond), the collagen capsule is complete. No contrast diffusion into the lucent center is seen, even with delayed imaging.20
Nonenhanced CT imaging is of limited value in the diagnosis of brain abscesses. In the setting of a large mass lesion, ventricular compression, mass effect, and edema may be observed (Fig. 1). When observed, such findings are suggestive of a brain abscess and should prompt further imaging. The presence of intracranial air on CT, another rare finding, may suggest a sinus or surgical source of infection or an anaerobic process.
Contrast-enhanced CT demonstrates a decrease in the attenuation of the center of the lesion. As the abscess matures, rim enhancement becomes prominent. A normal contrast-enhanced CT of the brain does not exclude an early cerebritis. Although contrast-enhanced CT scan is sensitive for the detection of brain abscesses, other processes, such as necrotic or metastatic tumors, resolving hematomas, and infarcts can have a similar appearance.21,22
Magnetic resonance imaging of the brain is more sensitive than contrast-enhanced CT scan in the detection of early cerebritis, brainstem lesions, ventricular or subarachnoid hemorrhage, associated satellite lesions, extent of central necrosis, ring enhancement, and cerebral edema.23,24 With T2-weighted imaging, the abscess and associated edema seem hyperintense relative to the surrounding normal brain tissue. Classically, abscesses are characterized by a hypointense capsule on T2-weighted images (Fig. 2).25 On T1-weighted imaging, the lesion and surrounding vasogenic edema are hypointense relative to the surrounding normal brain tissue. Marked enhancement of the capsule occurs with the administration of intravenous gadolinium (Fig. 3).
The center of an abscess is a liquefied core containing a complex protein matrix that binds water molecules, inflammatory cells, necrotic debris, and viscous pus. This matrix restricts translational movement of water. Consequently, diffusion-weighted imaging (DWI) demonstrates hyperintense signal within the center of the lesion, whereas the corresponding apparent diffusion coefficient shows low signal, indicating restricted diffusion (Figs. 4A, B).26 With rare exception, DWI is considered highly specific for brain abscess diagnosis.27,28
Radiation damage to the temporal lobe after therapy for nasopharyngeal and squamous cell carcinoma can result in clinical symptoms and radiological abnormalities that may mimic abscess formation. Delayed radiation necrosis should be considered when any patient who has received cranial irradiation develops focal neurological symptoms.29 Necrotic tumors may also have a similar appearance. However, the ring-enhancing lesion seen in radiation necrosis or necrotic tumor should not demonstrate the restricted diffusion seen with an abscess.30,31 In patients without a clear diagnosis after DWI, magnetic resonance spectroscopy or a tagged white blood cell scan may be helpful.32,33
Patients suspected of brain abscess should undergo evaluation for human immunodeficiency virus and other immunocompromised conditions. In individuals infected with human immunodeficiency virus, central nervous system (CNS) toxoplasmosis, primary CNS lymphoma, and progressive multifocal leukoencephalopathy are the most common causes for focal radiographic abnormalities of the brain. Features suggestive of toxoplasmosis include multiple or subcortical lesions, edema, and abscess wall thickness of less than 3 mm.30 Significant immunodeficiency also increases the likelihood of nonbacterial pathogens such as Nocardia species, fungal pathogens, and Mycobacterium tuberculosis.
As previously noted, lumbar puncture is not performed routinely in patients with brain abscesses, given the potential for brain herniation. In one series, 5 of the 22 patients who underwent lumbar puncture developed evidence of midbrain compression within 2 hours of the procedure.18 In another series, 8 of the 22 patients acutely deteriorated after lumbar puncture.5 As a consequence, lumbar puncture is usually contraindicated.
Ultimately, only biopsy or drainage can provide a definitive diagnosis of brain abscess and exclude a necrotic tumor. In most instances, surgeons prefer to perform central aspiration and drainage if abscess is suspected, although an excisional biopsy is preferred if a necrotic tumor is more likely.22,26 Unfortunately, aspirated abscess cultures may still remain sterile in most cases.8,11
TREATMENT AND OUTCOMES
Therapy of brain abscesses involves both surgical and nonsurgical interventions. Blood cultures should be obtained before the initiation of antibiotics because they may identify an etiology in 11% of the hematogenous cases.6 Empiric antibiotic therapy is indicated if the patient seems acutely ill or experiences an acute neurological deterioration. When intracranial extension of local infection is suspected, as in the current case, microaerophilic streptococci and anaerobes may be targeted with high-dose intravenous penicillin (10-20 million units per day) and metronidazole. The addition of a third-generation cephalosporin for Haemophilus species or rare facultative gram-negative anaerobes should also be considered.3 If intracerebral abscess is secondary to chronic otitis media, broader gram-negative coverage for Enterobacteriaceae and Pseudomonas is prudent. Antecedent head trauma or neurosurgery should prompt the initiation of therapy against S. aureus.
Surgical excision has largely been replaced by stereotactic abscess aspiration. This minimally invasive drainage procedure carries a morbidity rate of 0% to 1%.15 Risk factors for treatment failure after stereotactic needle aspiration include inadequate drainage, lack of catheter placement for larger abscesses, chronic immunosuppression, and insufficient antibiotic therapy.34 In select circumstances, brain abscesses may be managed with antibiotic therapy alone. The response to antibiotic therapy alone is greater in patients with small lesions (<2 cm) located in well-vascularized cortical areas.15 Rosenblum et al35 suggest that a nonoperative approach should be pursued in patients who are clinically stable and have concomitant meningitis, hydrocephalus requiring a shunt, deep lesions, or multiple distal lesions. Complete excision is still recommended for multiloculated abscesses or infection with more resilient pathogens, such as fungi or Nocardia species.3
Antimicrobial penetration into brain abscesses can be unreliable and often differs from antimicrobial diffusion across the blood-brain barrier. Certain antimicrobials, including third-generation cephalosporins, clindamycin, chloramphenicol, and metronidazole, have displayed adequate brain abscess penetration.14 Limited data on the use of vancomycin suggest that prolonged therapy may allow adequate abscess fluid drug concentrations.36 The fluoroquinolones and imipenem have been used to treat resistant pathogens with considerable success; however, these agents may result in a lower seizure threshold and must be used with caution.3 A small series of patients treated with surgical excision suggest that operative intervention may be followed with as little as 3 to 4 weeks of parenteral antibiotics.37 However, most authors support the use of 6 to 8 weeks of parental antibiotic therapy because the results of this study cannot be extrapolated to patients subjected to abscess aspiration or medical therapy alone. Intracavitary antibiotic therapy has been used in select cases of resistant or fungal infection, but current and less invasive therapeutic options have made this measure unnecessary in most circumstances.38
Repeat imaging with CT and MRI may not reveal improvement of mature abscess rim enhancement for 5 weeks or longer after antibiotic initiation or aspiration.39 One study suggests that repeat DWI of the brain abscess is more effective in determining a response to therapy.40
Adjunctive steroid therapy is often used in cases of brain abscess, although the use of this measure is unclear.5,6,16 In one experimental animal study, the use of steroids did not affect mortality, abscess size, or inflammatory response; however, it did delay abscess wall collagen deposition.41 Steroids have also been associated with impaired benzylpenicillin penetration in animal models of brain abscess.42 Despite these factors, most authors conclude that steroids may be beneficial in instances of increased intracranial pressure or severe edema.3 Because of the high risk of seizures, anticonvulsant therapy is also recommended in these cases.43
Mortality caused by bacterial brain abscess has declined from 30% to 60% in the 1970s to less than 21% in the present.5,16,44 This reduction in mortality has been attributed to the routine use of brain imaging in abscess diagnosis and aspiration.16 Patients with nonotogenic infections seem to fare worse; Yen et al8 report a mortality rate of 3.8% in patients with an otolaryngologic source, whereas the mortality rate in patients with brain abscesses of other etiology was 24%. A higher morbidity rate has been demonstrated in cases of gram-negative infection, low Glascow Coma Scale score, concomitant meningitis, deep-seated infection, or underlying immunodeficiency.1,4,15,16,44 Morbidity associated with brain abscesses can be substantial; more than 50% of the affected patients have neurological sequelae at the time of discharge.6,12 One study suggests that nearly 70% brain abscess patients seen in long-term follow-up experienced seizures.45
Intraventricular rupture of brain abscess (IVROBA), a rare complication of brain abscess, is associated with a high mortality rate.4,12,23,44 IVROBA is more common in patients with a hematogenous source of infection and in patients with deep-seated abscesses.44,46 IVROBA should be considered in any patient with acute decompensation during therapy.
Our patient received an 8-week course of intravenous antibiotics. One month after completion of therapy, the patient was seen in follow-up by a neurosurgeon. Magnetic resonance spectroscopy demonstrated resolution of abscess and also findings consistent with radiation necrosis. Consequently, the patient opted for palliative management.
1. Prasad KN, Mishra AM, Gupta D, et al. Analysis of microbial etiology and mortality in patients with brain abscess. J Infect
2. Osenbach RK, Loftus CM. Diagnosis and management of brain abscess. Neurosurg Clin N Am
3. Mathisen GE, Johnson JP. Brain abscess. Clin Infect Dis
4. Xiao F, Tseng MY, Teng LJ, et al. Brain abscess: clinical experience and analysis of prognostic factors. Surg Neurol
5. Yang SY. Brain abscess: a review of 400 cases. J Neurosurg
6. Chun CH, Johnson JD, Hofstetter M, et al. Brain abscess: a study of 45 consecutive cases. Medicine
7. Yang SY, Zhao CS. Review of 140 patients with brain abscess. Surg Neurol
8. Yen PT, Chan ST, Huang TS. Brain abscess: with special reference to otolaryngologic sources of infection. Otolaryngol Head Neck Surg
9. Calfee DP, Wispelwey B. Brain abscess. Semin Neurol
10. Le Moal G, Landron C, Grollier G, et al. Characteristics of brain abscess with isolation of anaerobic bacteria. Scand J Infect Dis
11. De Louvois J, Gortavai P, Hurley R. Bacteriology of abscess of the central nervous system: a multicenter prospective study. BMJ
12. Schliamser SE, Backman K, Norrby SR. Intracranial abscesses in adults: an analysis of 54 consecutive cases. Scand J Infect Dis
13. Mampalam TJ, Rosenblum ML. Trends in the management of bacterial brain abscesses: a review of 102 cases over 17 years. Neurosurgery
14. Yogev R, Bar-Meir M. Management of brain abscesses in children. Pediatr Infect Dis J
. Feb 2004;23(2):157-159.
15. Mamelak AN, Mampalam TJ, Obana WG, et al. Improved management of multiple brain abscesses: a combined surgical and medical approach. Neurosurgery
16. Seydoux C, Francioli P. Bacterial brain abscesses: factors influencing mortality and sequelae. Clin Infect Dis
17. Domingo Z. Clostridial brain abscesses. Br J Neurosurg
18. Samson DS, Clark K. A current review of brain abscess. Am J Med
19. Rosenblum ML, Hoff JT, Norman D, et al. Decreased mortality from brain abscesses since advent of computerized tomography. J Neurosurg
20. Britt RH, Enzmann DR, Yeager AS. Neuropathological and computerized tomographic findings in experimental brain abscess. J Neurosurg
21. Stevens JM, Hall-Craggs MA, Chong WK, et al. Cranial and intracranial pathology (2): infections; AIDS, inflammatory, demyelinating and metabolic diseases. In: Grainger and Allison's Diagnostic Radiology: A Textbook of Medical Imaging
. 4th ed. New York, NY: Churchill Livingstone; 2001;2377-2381.
22. Kastrup O, Wanke I, Maschke M. Neuroimaging of infections. NeuroRx
23. Walshe P, Brennan P, Walsh M, et al. Computerized tomography is not reliable in the diagnosis of brainstem infection. J Laryngol Otol
24. Kamra P, Vatsal DK, Husain M, et al. MRI demonstration of unsuspected intraventricular rupture of pyogenic cerebral abscesses in patients being treated for meningitis. Neuroradiology
25. Haimes AB, Zimmerman RD, Morgello S, et al. MR imaging of brain abscesses. AJR Am J Roentgenol
26. Leuthardt EC, Wippold FJ 2nd, Oswood MC, et al. Diffusion-weighted MR imaging in the preoperative assessment of brain abscesses. Surg Neurol
27. Bukte Y, Paksoy Y, Genc E, et al. Role of diffusion-weighted MR in differential diagnosis of intracranial cystic lesions. Clin Radiol
28. Tung GA, Evangelista P, Rogg JM, et al. Diffusion-weighted MR imaging of rim-enhancing brain masses: is markedly decreased water diffusion specific for brain abscess? AJR Am J Roentgenol
29. Morris JG, Gratten-Smith P, Panegyres PK, et al. Delayed cerebral radiation necrosis. Q J Nucl Med
30. Wong J, Quint DJ. Imaging of central nervous system infections. Semin Roentgenol
31. Chang SC, Lai PH, Chen WL, et al. Diffusion-weighted MRI features of brain abscess and cystic or necrotic brain tumors: comparison with conventional MRI. Clin Imaging
32. Liberatore M, Drudi FM, Tarantino R, et al. Tc-99m exametazime-labeled leukocyte scans in the study of infections in skull neurosurgery. Clin Nucl Med
33. Dev R, Gupta RK, Poptani H, et al. Role of in vivo proton magnetic resonance spectroscopy in the diagnosis and management of brain abscesses. Neurosurgery
34. Kondziolka D, Duma CM, Lunsford LD. Factors that enhance the likelihood of successful stereotactic treatment of brain abscesses. Acta Neurochir (Wien)
35. Rosenblum ML, Hoff JL, Norman D, et al. Nonoperative treatment of brain abscesses in selected high-risk patients. J Neurosurg
36. Levy RM, Gutin PH, Baskin DS, et al. Vancomycin penetration of a brain abscess: case report and review of the literature. Neurosurgery
37. Sjolin J, Lilja A, Eriksson N, et al. Treatment of brain abscess with cefotaxime and metronidazole: prospective study on 15 consecutive patients. Clin Infect Dis
38. Mahmood A, Abad RM, Updegrove JH. Brain abscess treated by continuous antibiotic perfusion: technical note. Neurol Res
39. Chen SC, Chung HW. Diffusion-weighted imaging parameters to track success of pyogenic brain abscess therapy. AJNR Am J Neuroradiol
40. Cartes-Zumelzu FW, Stavrou I, Castillo M, et al. Diffusion-weighted imaging in the assessment of brain abscesses therapy. AJNR Am J Neuroradiol
41. Schroeder KA, McKeever PE, Schaberg DR, et al. Effect of dexamethasone on experimental brain abscess. J Neurosurg
42. Kourtopoulos H, Holm SE, Norrby SR. The influence of steroids on the penetration of antibiotics into brain tissue and brain abscesses. An experimental study in rats. J Antimicrob Chemother
43. Shaw MDW. Post-operative epilepsy and the efficacy of anticonvulsant therapy. Acta Neurochir (Wien)
44. Takeshita M, Kagawa M, Izawa M, et al. Current treatment strategies and factors influencing outcomes in patients with bacterial brain abscess. Acta Neurochir (Wien)
45. Legg NJ, Gupta PC, Scott DF. Epilepsy following cerebral abscess: a clinical and EEG study of 70 patients. Brain
46. Waggener JD. The pathophysiology of bacterial meningitis and cerebral abscesses: an anatomical interpretation. Adv Neurol
© 2006 Lippincott Williams & Wilkins, Inc.