Despite the many advances we have made in imaging techniques, laboratory diagnosis, antibiotic therapy, and surgery, brain abscesses remain a challenging clinical problem with a high morbidity and mortality rate. Brain abscesses can be caused by bacteria, mycobacteria, fungi, and parasites (protozoa and helminths), with reported incidence rates ranging from 0.4 to 0.9 cases per 100,000 people. The incidence is increased in immunocompromised patients.
A recent NEJM review was published summarizing the pathogenesis, clinical presentation, recommended antibiotic therapy, advances in surgical treatment, and clinical prognosis of brain abscesses.
Pathogenesis and epidemiology
In the majority of patients, brain abscesses are due to susceptibility factors such as underlying disease (e.g., history of HIV infection), treatment with immunosuppressive drugs, disruption of the natural protective barrier around the brain (e.g., trauma due to surgery, trauma, mastoiditis, sinusitis, or oral infection) or systemic infection (e.g., endocarditis or bacteremia). In half of the patients, the bacteria enter the brain by diffusion through the adjacent tissues, and in 1/3 of the cases by blood-borne transmission (Figure 1).
The pathogenesis of the infection depends on the susceptibility status of the organism. Patients who are immunocompromised due to treatment with immunosuppressive agents as a result of solid organ or hematopoietic stem cell transplantation usually present with tuberculosis or nonbacterial infections such as fungal or parasitic infections.
Brain abscesses in HIV-infected patients are usually due to Toxoplasma gondii infection, but HIV can also predispose patients to Mycobacterium tuberculosis infection. Patients who have received solid organ transplants are not only susceptible to Nocardia brain abscesses, but also to fungal infections, such as Aspergillus or Candida infections. Ninety percent of these patients have fungal infections.
Brain abscesses can also develop after neurosurgery or after head trauma. Infections in these patients are usually caused by skin surface bacteria such as Staphylococcus aureus, Staphylococcus epidermidis, or Gram-negative bacilli. Brain abscesses due to spread from adjacent tissues (e.g., otitis media, sinusitis, mastoiditis) are usually caused by streptococci, but can also present as staphylococcal abscesses and multibacterial infectious abscesses (both anaerobic and gram-negative bacilli).
Blood-borne transmission of bacteria is usually associated with underlying cardiac disease (e.g., endocarditis or congenital heart disease), pulmonary disease (e.g., arteriovenous fistula), or infection of distant lesions (skin, paranasal sinuses, and teeth). Blood-borne brain abscesses are mainly caused by staphylococci and streptococci. Blood-borne infections caused by paranasal sinuses and teeth are usually polymicrobial in nature.
The first stage of a brain abscess usually presents as early encephalitis, caused by an inflammatory response around the necrotic center and increased edema of the surrounding white matter. Subsequently, the necrotic center reaches its maximum volume and produces a cystic cavity through mechanisms such as fibroblast proliferation and neovascularization. The cystic cavity thickens with increased active collagen, but the size of the inflammation and edema exceeds the size of the cystic cavity. Figure 2 shows the pathological presentation of a brain abscess.
Clinical manifestations
The most common clinical manifestation of brain abscess is headache, and rarely fever and altered level of consciousness are seen. Neurologic signs depend on the location of the abscess lesion and may be only mild for the first few days or weeks of the abscess. Patients with brain abscesses in the frontal or right temporal lobe may present with behavioral changes. Brain abscesses in the brainstem and cerebellar areas may present with cranial nerve palsy, gait disturbances, headaches (due to hydrocephalus), or altered state of consciousness.
Seizures may manifest in 25% of patients. The clinical manifestations become more pronounced as the abscess enlarges and the edema around the lesion increases. However, these signs and symptoms may be difficult to recognize because of the use of sedative medications or the effects of underlying neurologic disease. Patients with hematologically transmitted brain abscesses will present with manifestations of the original lesion infection.
The differential diagnosis of brain abscess includes a range of neurologic and infectious brain diseases such as tumors, stroke, bacterial meningitis, epidural abscesses, and subdural pus. patients with HIV infection also need to be considered for the differential of primary central nervous system lymphoma.
Diagnosis
Cephalometric imaging should be performed in all patients with suspected brain abscesses. Enhanced CT is a rapid imaging method to detect the size, number and location of abscesses. Magnetic resonance imaging (MRI) combined with DWI,ADC sequences is a valuable diagnostic tool for differentiating brain cysts from primary tumors, cystic tumors, and tumor necrosis (Figure 3).
In a prospective study that included 147 cystic lesions in 115 patients, 97 of which were brain abscesses, DWI imaging had high sensitivity and specificity (98% positive predictive value and 92% negative predictive value) for differentiating brain abscesses from primary tumors or metastatic carcinomas. Proton nuclear magnetic resonance (1hNMR) spectral imaging can also be used for differential diagnosis, but its specificity and sensitivity in combination with DWI is only slightly increased compared with DWI alone.
Blood and cerebrospinal fluid cultures are required in approximately 1/4 of patients to identify the causative pathogen. Patients with combined meningitis may be more amenable to lumbar cerebrospinal fluid examination. However, the risk of brain herniation needs to be considered in these patients as well.
Lumbar puncture should only be considered if there is clinical suspicion of meningitis or abscess breaking into the ventricular system and there are no contraindications to lumbar puncture (imaging showing hemispheric displacement or coagulopathy). Cultures should also be performed for potential infections of the teeth, paranasal sinuses, ears, and skin areas, and surgical treatment may also be required to remove the infected focus.
Surgical treatment
If the pathogen is unknown in a patient with a brain abscess, neurosurgery may be performed in select patients to identify the pathogen and reduce the size of the abscess lesion. If modern stereotactic neurosurgical techniques are used, almost all brain abscesses greater than or equal to 1 cm in diameter can be subjected to stereotactic suction surgery, regardless of their location.
Stereotactic navigation systems can be used for abscess drainage, and volumetric CT or MRI imaging techniques can be used for three-dimensional reconstruction of the patient’s brain. Careful trajectory planning is then required to select the most optimal puncture site to avoid damaging useful brain areas (e.g., those responsible for language, motor, sensory, visual, and other functions).
Stereotactic surgery in the center of the septic lesion should be performed for both diagnostic and decompression purposes, unless the type of infecting agent or the patient’s condition does not allow it. If cranial imaging does not show the formation of an abscess cavity, careful consideration and choice between performing a stereotactic biopsy and empiric antibiotic therapy is required.
In HIV patients with probable Toxoplasma gondii infection, presumptive antimicrobial therapy may be given if the anti-Toxoplasma gondii IgG is positive but a histologic diagnosis is lacking. In rare cases, surgical treatment may not be an option when the patient is in poor health or when co-morbidities would increase the risk of surgery. If stereotactic techniques are not possible, direct abscess drainage can be performed by transcranial ultrasound through a hole or small debridement flap, but this approach is not recommended for small abscesses deep in the brain.
Diagnostic suctioning is aimed at maximizing abscess drainage. Continuous drainage through a catheter connected to the abscess cavity is used as a method that may reduce the probability of reoperation, but is not routinely recommended. Some experts recommend using this drainage catheter postoperatively to allow antibiotics to enter the abscess cavity, as systemic antibiotic therapy has limited access to the abscess cavity, but it is also not routinely recommended due to the relatively small amount of data on the benefits and risks of this approach.
Complete resection has been recommended since more than 20 years ago, but its role is now seen as limited due to the development of medical therapy and minimally invasive neurosurgical treatment. However, excisional treatment may be considered instead of drainage if the abscess is superficial and not located in a significant functional brain area, especially if fungal or tuberculous infection or branching mycobacteria (e.g., Actinomyces or Nocardia spp.) are suspected.
If the causative pathogen is identified, the indication for abscess aspiration depends on the size and location of the lesion, the patient’s clinical condition, and the likelihood of achieving meaningful decompression by aspiration. In a small case series, treatment was prone to failure when treated with antibiotic medication alone.
Neurosurgical intervention was recommended for abscess lesions larger than 2.5 cm in diameter. However, due to the lack of data from comparative studies, lesion size cannot be used as an absolute indication for attraction surgery. In patients with multiple small abscess foci, aspiration should be performed on the largest of these foci to clarify the diagnosis; whether aspiration is performed on the others depends on the size of the lesion, the surrounding edema, the patient’s symptoms, and the response to antibiotic therapy.
For abscesses that have led to hemispheric displacement and may cause brain herniation, this may suggest the need for neurosurgical intervention, regardless of the size of the abscess. For abscesses that are adjacent to the ventricular system but have not yet ruptured into the ventricles, drainage procedures may be considered to prevent rupture of the abscess or result in inflammation of the ventricles.
Microbiologic evaluation of cerebrospinal fluid, blood, or abscess drainage should include Gram staining and culture under aerobic and anaerobic conditions. In immunocompromised and high-risk patients (e.g., those with a history of tuberculosis or opportunistic infections), cultures for Mycobacterium, Nocardia spp. and fungi should be performed, along with PCR testing for Toxoplasma gondii. If bacterial brain abscess is strongly suspected but culture results are negative, PCR16s ribosomal DNA sequencing is feasible and can clarify the diagnosis and guide the next step of antibiotic therapy.
A study showed that when suction drains from 71 brain abscess patients were tested, only 30 patients had positive bacterial cultures and 59 patients tested positive for bacterial DNA. The investigators identified 80 different bacterial taxa, 44 of which had not been present in previous pathogens of brain abscesses, including 37 that had not been present in previously reported bacterial cultures. Although these data suggest a high degree of bacterial variability in brain abscesses, it is unclear whether these genera are involved in brain abscess development and whether treatment is required.
Antibiotic therapy
Delayed initiation of antibiotic therapy may lead to a poor prognosis, as shown in a retrospective study in which the mean time interval between definitive diagnosis and initiation of antibiotic therapy was approximately 2 days. The researchers concluded that antibiotic therapy should be administered as soon as a brain abscess is clinically suspected.
Because antibiotic treatment before stereotactic aspiration of the abscess may reduce the likelihood of diagnostic testing of cerebrospinal fluid, it is reasonable to delay antibiotic treatment until after neurosurgical intervention, provided that the condition is less severe, the patient is clinically stable, and surgical treatment can be completed in a short period of time. However, caution is needed in adopting this strategy because abscesses can progress rapidly and at an unexpected rate, regardless of the initial severity of the disease.
The starting antibiotic therapy should be selected for the pathogenic organism most likely to cause the disease, taking into account the mechanism of infection, the patient’s previous susceptibility status, the type of antibiotic therapy to which the patient is susceptible, and the ability of the antibiotic to penetrate the abscess (Tables 1 and 2).
Patients after organ transplantation should receive empiric antibiotic therapy, such as triple cephalosporins (ceftriaxone or cefotaxime) plus metronidazole for bacterial brain abscesses, cotrimoxazole or sulfadiazine for Nocardia spp. infections, and voriconazole for fungal infections, especially Aspergillus infections.
For the initial treatment of HIV-infected patients, the addition of a therapeutic agent against Toxoplasma gondii (etanercept + sulfadiazine) is recommended, but only for patients with positive IgG antibodies to Toxoplasma gondii. Drug therapy targeting TB (isoniazid, rifampin, pyrazinamide, and ethambutol) should be considered for patients with HIV infection or who have traveled to areas and countries where TB is endemic, or who have known risk factors for TB.
For patients with post-neurosurgical or cranial trauma with fractures, their empiric treatment medications include vancomycin plus third- or fourth-generation cephalosporins (i.e., cefepime) and metronidazole. In patients with an extracranial source of lesion and no history of neurosurgical treatment, treatment with ceftriaxone or cefotaxime in combination with metronidazole should be used.
Vancomycin may be added if staphylococcal infection is suspected. Meropenem may be used in patients with contraindications to cephalosporin or metronidazole therapy. A retrospective Spanish study showed that patients treated with cefotaxime + metronidazole and those treated with meropenem had a similar prognosis.
For patients with blood-borne brain abscesses, therapeutic agents include triple cephalosporins in combination with metronidazole for covering anaerobes, plus vancomycin for possible staphylococcal infections, depending on the results of microbiological testing as well as the results of in vitro susceptibility testing.
Once the pathogen of the infection is identified, antibiotics are the most effective treatment (Table 2). A dilemma arises when blood cultures show a single pathogen infection. Because 27% of brain abscesses are multibacterial, treatment with broad-spectrum antibiotics is recommended until the abscess culture is clear, or aerobic and anaerobic cultures show no other pathogenic infection. However, if the infection originates from an adjacent lesion, broad-spectrum antimicrobial therapy should be used to cover multiple pathogens (including anaerobes) even if no other pathogens have been isolated.
Multi-resistant Gram-negative bacterial infections with brain abscesses have been reported after neurosurgery and following complex cranial trauma. Fungal brain abscesses respond poorly to antibiotic therapy, even though one study showed a reduction in mortality with voriconazole treatment (65% vs 91% historical control).
The duration of intravenous antibiotic therapy in patients with bacterial brain abscess is traditionally considered to be 6-8 weeks. Prolonged treatment with metronidazole may be associated with the development of neuropathy. However, in one study, peripheral neuropathy was shown to improve after cessation of metronidazole treatment.
The British Association for Antibiotic Therapy Neurosurgical Infection Working Group recommends a 1-2 week duration of intravenous antibiotic therapy for patients with bacterial brain abscess, after which the type of antibiotic therapy should be reasonably altered and adjusted according to clinical response. This approach has been successfully applied in a select group of patients who are elective, but not as standard of care. Oral antibiotic therapy in these patients includes metronidazole, ciprofloxacin, and amoxicillin.
Important criteria for assessing treatment are the patient’s neurological symptoms and the size of the abscess as demonstrated by cranial imaging. If clinical deterioration is present cranial imaging should be performed immediately. If there is no improvement, a repeat examination should be performed in 1-2 weeks and repeated every 2 weeks for 3 months thereafter until clinical resolution is achieved. Further neurosurgical procedures are indicated when cranial imaging shows signs of lesion enlargement and clinical deterioration despite antibiotic therapy.
Complications and prognosis
If the patient has a decreased level of consciousness, immediate cranial imaging is required to check for the presence of hydrocephalus or impending brain herniation. Rupture of an abscess into the ventricular system causes ventriculitis, leading to hydrocephalus, which is associated with a high mortality rate (27%-85%). In patients with ruptured abscesses, placement of a ventricular catheter provides ventricular drainage, aspiration of cerebrospinal fluid for examination, monitoring of intracranial pressure, and a direct intracerebroventricular route for administration of antibiotic therapy.
Hydrocephalus is a common complication in patients with posterior cranial fossa abscesses. Decreased level of consciousness can be due to seizures and persistent epilepsy. There are no randomized studies on the prophylactic use of antiepileptic drugs in patients with brain abscesses. In a study of patients with brain tumors, prophylactic treatment with antiepileptic drugs did not reduce the frequency of seizures. Routine antiepileptic therapy is not recommended for patients with brain abscesses.
Patients may have increased neurological deficits as the abscess increases in size and as peripheral edema worsens. Adjunctive treatment with glucocorticoids may reduce cerebral edema in about half of patients treated with this therapy.
However, because of the lack of data from randomized studies and the fact that glucocorticoids reduce the entry of antimicrobial drugs into the central nervous system, their use should be limited in patients with significant cerebral edema who are at risk for brain herniation. Only a few small case reports mention the use of hyperbaric oxygen as an adjunctive treatment and not as a routine treatment.
The prognosis of patients with brain abscesses has improved considerably over the past 50 years due to improvements in brain imaging techniques, increased use of antibiotic therapy, and the introduction of minimally invasive neurosurgical treatment. The mortality rate has decreased from 40% in 1960 to 15% today. Currently, 70% of patients with brain abscesses have a good prognosis, with no neurological sequelae or very mild symptoms.