Abstract
We report three cases of brain abscess in children admitted to QECH in 2006. All children were HIV-uninfected. One case was associated with staphylococcal empyema, another with chronic suppurative otitis media and mastoiditis, and the third case had no identified extracranial focus of infection. These cases illustrate the difficulties of diagnosis and management of brain abscesses in the resource-poor setting where other causes of infection of the central nervous system are common. The typical clinical presentation of brain abscess of altered mental state and seizures is also characteristic of cerebral malaria and meningitis and it is likely that many cases of brain abscess in Malawian children are not diagnosed. The value of cranial CT scan, ideally with contrast, for diagnosis and management of brain abscess is highlighted by these cases.
Introduction
Brain abscess is relatively uncommon but a serious disease among children. Recent experience of tertiary paediatric hospitals in Australia and Pakistan report an average of three cases per year.1,2 The incidence in Malawi is uncertain because diagnosis is difficult where neuroradiological imaging is usually unavailable and there is clinical overlap with more common forms of central nervous system (CNS) infection. For example in the malaria-endemic setting, children presenting with altered mental state and seizures may be misdiagnosed as cerebral malaria.3 The installation of a computerized tomography (CT) scanner at the Queen Elizabeth Central Hospital (QECH), Blantyre, has improved diagnosis and management of acute encephalopathy in children. We report three cases of suppurative intracranial abscess in HIV-uninfected children presenting between January and July 2006.
Case reports
Case 1
A 2-year 7-month old boy presented with a two-week history of fever, cough and lethargy, and worsening dyspnoea for two days. He had been admitted a week earlier to a peripheral hospital for four days with the same complaints but he continued to deteriorate and so his mother discharged him and brought him to QECH for assessment. Management at the peripheral hospital was unclear. There was no history of headache, seizure or vomiting. His past medical history revealed a previously well, fully immunized child with no known contact with tuberculosis (TB).
Clinical examination revealed a well-nourished boy (weight 12 kg; weight-for-height >85%). He was comatose with a nonlocalising withdrawal response to painful stimuli only. He was pale, febrile (temperature 38.3°C), tachycardic (pulse rate 160 bpm) with a normal blood pressure (80/60 mmHg). He had severe respiratory distress with tachypnoea (respiratory rate 60 bpm), grunting, nasal flaring and subcostal indrawing. He was not cyanosed but was hypoxic with oxygen saturation of 79% in air. Chest examination revealed reduced air entry and bronchial breathing on the right side. The abdominal examination revealed hepatosplenomegaly of 6 cm each. The rest of the examination was unremarkable. There were no focal neurological signs. Pupillary size and reaction were normal as was fundal examination. Cardiac examination was normal.
Initial investigations revealed hypoglycaemia (blood glucose 2.6 mmol/L: normal 3.5–6.5) corrected with a bolus of intravenous dextrose, and elevated serum lactate (15.2 mmol/L: normal < 4.0). A thick blood film for malaria showed 1 – 10 asexual Plasmodium falciparum per high-powered film and packed cell volume (PCV) was 13%. Lumbar puncture was traumatic and the cerebral spinal fluid (CSF) showed a white cell count of 5/mm3, red cell count of 1600/mm3, protein 30mg/dL and glucose 99 mg/dL (normal 60–120). Blood culture was performed and a chest X-ray (CXR) was planned for when stable.
The initial diagnoses were pneumonia and severe malarial anaemia. He was commenced on chloramphenicol 25mg/kg, a loading dose of quinine (20mg/kg), oxygen by nasal prongs and was transfused with whole blood (20ml/kg) an hour later. Five hours later the patient deteriorated. He became unresponsive with lack of withdrawal to painful stimuli, brisk deep tendon reflexes, a fixed upward gaze, acidotic breathing and hypoglycaemia. He was given dextrose and parenteral gentamicin (6mg/kg/day) was added. There was considerable improvement by day 3. He was fully conscious and eating, PCV was 26% with no malaria parasitaemia on peripheral blood film. HIV test was negative. CSF culture result was no growth. He was still febrile (axillary temperature 38°C) with chest signs consistent with pneumonia and a CXR on day 4 showed diffuse opacification of the right lung. Chloramphenicol was stopped and cefuroxime commenced. Staphylococcus aureus resistant to penicillin and chloramphenicol but sensitive to gentamicin, erythromycin, tetracycline and cotrimoxazole was isolated from blood culture on day 4. He was still sick, febrile with increasing respiratory distress and had developed dullness on percussion over the right chest. Chest ultrasonography showed a 4 cm effusion in the right chest and pleural aspirate revealed an empyema. Cloxacillin was added. The surgeons were consulted and he was scheduled for an intrathoracic chest drain the following morning.
But that night his condition deteriorated. He had repeated prolonged generalized seizures and became comatose. A random blood sugar was normal. Oxygen and paraldehyde were administered with cessation of seizures. Ceftriaxone (100 mg/kg/day) was added to the cloxacillin. A chest drain was inserted which drained 400ml of thick pus. The pleural fluid also grew Staphylococcal aureus.
Over the next 3 weeks, the patient remained comatose with only a withdrawal response to painful stimuli. The possibility of an intracranial abscess was considered or of encephalopathy secondary to septicaemia and hypoxia. A cranial CT scan revealed a lesion in the left parietal-temporal region (Figure 1) reported as consistent with a haemorrhage. He continued to remain febrile and toxic. Surgeons were consulted and advised against surgery at this stage especially as he still had major respiratory problems. He developed decubitus ulcers and a broncho-pleural fistula with a persistent collapsed right lung despite effective drainage. Recurrent chest drain obstruction required four replacements. On day 24, he required another blood transfusion. On day 30, his fever began to settle. He became more responsive and was able to feed orally but had right-sided weakness. An ophthalmologist assessment on day 32 revealed marked visual loss and sluggish pupils but normal fundi. A second CT scan with contrast showed a lesion with enhanced margins consistent with a brain abscess (Figure 2). The treatment was continued and he was discharged on day 54 with 1 weeks' supply of cloxacillin having already received cloxacillin for 7 weeks. At follow-up one month later, his progress was excellent. He was walking normally, feeding well, talking and his eyes were able to fix and follow.
Figure 1.
Initial CT scan of case 1 without contrast
Figure 2.
Follow up CT scan with contrast one month after treatment for case 1
Figure 1b.
Initial CT scan od case 1 with contrast
Case 2
6 year-old girl presented with a history of fever, headache and neck stiffness for 3 days and generalized seizures for a day. She was fully immunized including BCG immunization. The only remarkable history was of bilateral chronic suppurative otitis media for a year. She had no history of TB contact.
On examination, she was thin (weight 17 kg), febrile (axillary temperature 38.7°C), alert and well perfused. Her pulse was 136 bpm, respiratory rate of 34 bpm and BP of 100/70 mmHg. She had marked neck stiffness, a right lateral gaze preference with generalised symmetrical brisk reflexes. She had bilateral chronic suppurative otitis media and a tender left mastoid but no swelling of the mastoid. The rest of the examination was normal.
A lumbar puncture revealed cloudy CSF and blood film was negative for malarial parasites. Ceftriaxone (100mg/kg per day) was commenced. The blood glucose was 6.5 mmol/L and lactate 5 mmol/L. The CSF showed a white cell count of 15/mm3, no red blood cells, protein 100mg/dL and glucose of 252mg/dL. No organisms were seen. HIV test was negative. Blood culture was negative. On day 3 she started improving and her temperature settled. On day 5 she became irritable, confused and was unable to sit. A differential diagnosis of TB meningitis was considered. A chest X-ray revealed right perihilar infiltrations and tuberculin skin test (TST) was reactive with a blistering reaction of 20mm after 48 hours. She was commenced on TB meningitis treatment and prednisolone 2mg/kg daily on day 8. On day 10, a protruding left ear and a swelling over the left mastoid was noted. The diagnosis of intracranial abscess was entertained and a cranial CT scan revealed air and a lesion in the posterior fossa with midline shift of the cerebellum (Figure 3) and dilated ventricles. The surgeons were consulted and opted for a conservative approach with ceftriaxone, metronidazole (7.5mg/kg 8 hourly) and cloxacillin (25mg/kg 6 hourly). The TB meningitis treatment was discontinued. The patient responded and was discharged on day 29 on oral ciprofloxacin and flucloxacillin for 3 weeks.
Figure 3.
CT scan for case 2
Case 3
An 11-year old boy was referred from a district hospital by the medical staff, where he had been admitted for a week after presenting with headache and fever for 4 weeks. At the district hospital, he tested positive for malaria and was treated with intravenous quinine, benzylpenicillin and chloramphenicol for suspected bacterial meningitis. Four days later while in the district hospital, he developed a left sided hemiparesis with left sided focal seizures and was then referred to QECH for further management. At QECH, his past medical history revealed a previously well boy, fully immunized with no history of TB contact.
On examination, his temperature was 37.5C, blood pressure 115/80 mmHg, respiratory rate of 28/min and a pulse rate of 73/min. He was alert, well perfused, with neck stiffness, a positive Kernig's sign and a left sided hemiparesis. Direct ophthalmoscopy revealed bilateral disc ‘fullness’. The rest of the examination was unremarkable. A diagnosis of meningitis was made with possible differential diagnoses of intracranial abscess and TB meningitis also considered.
Lumbar puncture revealed clear CSF under low pressure. Intravenous ceftriaxone (100mg/kg/day) was commenced. The CSF showed white cell count of 18/mm3, red blood cell count 10/mm3, a protein 100mg/dL, glucose 252mg/dL and Gram stain negative. On day 2, he became stable, afebrile and CSF had no growth. Ceftriaxone was discontinued. HIV test was negative and TST was non-reactive after 72 hours. A cranial CT revealed a right subdural abscess measuring 3.5 × 2.5 cm with a midline shift. The ceftriaxone was restarted and surgical drainage was scheduled for the following day. On the same night, the patient developed focal left-sided seizures which responded to paraldehyde. A subdural empyema and haematoma were drained and metronidazole was added. His clinical condition improved becoming afebrile and fully conscious. On day 14, he was still complaining of a painful neck and pus was discharging from the incision site. Surgical wound exploration and debridement was done. Over the following 2 weeks, the boy was clinically stable except for recurrent pus discharge over the incision site which required surgical drainage and washouts on day 21, 23, 25 and 29. A repeat cranial CT on day 30 revealed multiple right intracerebral abscesses (the largest ones located in frontal and parietal regions) with oedema and shift of the midline to the left. At surgery, pus was drained from abscesses located in occipital and parietal regions but pus was not sent for Gram stain and culture. The surgical drains were removed after 48 hours. He remained stable over the next 15 days and on day 56, a repeat CT scan showed several small abscesses in the right cerebral hemisphere. Ceftriaxone and metronidazole were continued. By day 60, the facial palsy was improving, he was able to sit and eat. He was discharged on oral cefuroxime and oral metronidazole for two weeks. He was reported to have died at home a week later but no details were provided.
Discussion
Suppurative brain abscess is uncommon in children. Recent retrospective studies from tertiary referral hospitals in Pakistan and Australia reported 30 cases each over a ten-year period.1,2 Most cases occurred in children over 4 years, case-fatality was 10–16% and the majority of survivors had long-standing neurological deficit. A recent review states that 25% of all brain abscesses occur in children under the age of 15 years with a peak at 4–7 years and a high morbidity and mortality.4 Brain abscess is likely to be underreported from Malawi where diagnostic imaging is usually not available.
The pathogenesis of brain abscess was recently reviewed.4,5 Infection can arise either directly from a contiguous site such as middle ear, sinuses, teeth, and compound skull fractures or indirectly via haematogenous spread from extracranial sources such as pulmonary infection, endocarditis or septicaemia. Organisms involved in children include Streptococci (anaerobic and aerobic) responsible for about 50–70%, staphylococci (10–30%) and enteric bacteria (10–25%).4,6 Mixed flora can be found in up to 30%. The location and predisposing condition of the abscess influence aetiology.4,5 Abscesses arising from sinus or odontogenic foci tend to be frontal and caused by streptococci (aerobic and anaerobic), enterobacteriaceae, Staphylococcus aureus and anaerobes. Temporal or cerebellar lesions are usually otic in origin with mixed flora that include aerobic and anaerobic streptococci, Enterobacteriaceae and Pseudomonas aeruginosa. Abscesses caused by haematogenous spread are usually due to a single pathogen (Streptococcus milleri and Staphylococcus aureus are important causes) and occur in the distribution of the middle cerebral artery.3–5 A wider range of pathogens can cause brain abscess in immunocompromised children including fungi and mycobacteria.5,6
The three cases presented in this report were all previously well HIV-uninfected children. In two cases, a likely source of infection was identified - one with staphylococcal pulmonary infection and one with chronic suppurative otitis media and mastoiditis. The location of the abscesses was consistent with these associations. The one case with subdural empyema and multiple abscesses did not have an obvious extracranial source of infection.
A causative pathogen was not isolated from the abscess in any of the cases. It could be presumed in case 1 that the cause was Staphylococcal aureus as it was cultured from blood and pleural fluid. This was why cloxacillin was used for 8 weeks. Ceftriaxone was also used because of concerns that the brain abscess might be a complication of Salmonella septicaemia because at initial presentation case 1 had severe malarial anaemia as well as severe pneumonia. Severe malarial anaemia is commonly associated with bacteraemia due to non-typhoidal Salmonella with a risk of focal sepsis such as Salmonella meningitis or pneumonia following discharge.7,8
All three children had clinical features consistent with the presentation of a brain abscess4,5 at the time the diagnosis was considered - fever, seizures and altered mental state. It is possible that earlier drainage of the staphylococcal empyema might have reduced the risk of brain abscess in Case 1 although it is also possible that the acute but temporary neurological deterioration that occurred 5 hours after admission correlated with the beginning of infective cerebritis which later evolved to become an abscess. Case 1 was also complicated with multiple pathologies which may have contributed to the neurological deterioration on day 6 before the chest drain was inserted. He had had malaria, severe anaemia, staphylococcal septicaemia and hypoxia although had been stable for some days. TB meningitis was considered in Case 2 because of lack of response to what is usually very effective antibiotics for bacterial meningitis (i.e. ceftriaxone) and it is usual to consider that diagnosis in those circumstances especially with some leucocytes in the CSF and a reactive TST. The diagnosis of brain abscess was entertained when the mastoid swelling was noted and in retrospect, the finding of a tender (though not swollen) mastoid on admission should have been considered more carefully.
These cases highlight the value of having CT scan for more accurate diagnosis. The presentation of fever, seizures and altered mental state is very common in children admitted to QECH. The commonest causes are cerebral malaria and acute bacterial meningitis. A positive blood film for malaria does not necessarily mean that the cause of coma is cerebral malaria3 and a negative blood film does not necessarily exclude the diagnosis. The diagnosis of cerebral malaria is strongly supported by characteristic findings on examination of the fundus9 so that other diagnoses should be more strongly considered if malarial retinopathy is absent. However, malarial retinopathy may not always be present with cerebral malaria and its recognition requires some training and experience. Bacterial meningitis is usually a straightforward diagnosis as long as there is a high index of suspicion and microscopy of CSF is available. Once cerebral malaria and bacterial meningitis have been excluded or there is a poor response to treatment for the same, other diagnoses are usually considered including viral encephalitis, TB meningitis and brain abscess. Children with TB meningitis will usually have typical CSF findings of lymphocytosis and elevated protein while the CSF findings for brain abscess that has not ruptured into a CSF compartment are non-specific with overlap of findings with acute encephalitis. The three cases reported had only 5 to 18 white cells per mm3 in the CSF.
The possibility of clinical overlap with other conditions that require different management emphasizes the value of available imaging such as CT scan, preferably with contrast as abscesses show a typical ring enhancement. The use of contrast is not routine with cranial CT at QECH (and contrast is not always available) but ideally if brain abscess is considered, contrast would be used as it shows a characteristic enhancement of the abscess margins. Imaging is also required to direct surgical intervention which is an important part of the management.4,5 In carefully selected patients, surgical drainage is required in patients with illness for >2 weeks, neurologically impaired with signs for raised intracranial pressure and abscesses > 3 cm. Minimal invasive surgery with CT or MRI guided stereotactic aspiration have replaced traditional open craniotomies for aspiration in the developed world. Surgery can reduce the raised intracranial pressure and aid to obtaining pus for microbial diagnosis and enhance antibiotic efficacy.
Cranial ultrasound is another imaging technique that is much more readily available than CT scan but its use is limited to children with open fontanelles. Ultrasound is useful for diagnosis of subdural collections as they tend to be more common in children under 2 years of age.2 This diagnosis is not uncommon in our setting where the usual cause is non-typhoidal Salmonella (unpublished observation, SM Graham) and repeated needle aspiration through the fontanelle is part of management. The typical age range of brain abscess is in older children2,3 and so ultrasound is helpful for diagnosis.
The management of brain abscess in children is challenging. The existence of other diseases with similar signs and symptoms can delay the diagnosis. The most valuable tool in this setting is to have a high index of suspicion for brain abscess and to assess carefully for risk factors such as chronic ear disease or cardiac disease. Prompt referral to a centre with adequate support that includes prolonged broad-spectrum antibiotics and surgical drainage. The availability of neuroradiological imaging is extremely useful for accurate diagnosis and effective management.
References
- 1.Leotta N, Chaseling R, Duncan G, Isaacs D. Intracranial suppuration. J Pediatr Child Health. 2005;41:508–512. doi: 10.1111/j.1440-1754.2005.00693.x. [DOI] [PubMed] [Google Scholar]
- 2.Mehnaz A, Syed AU, Saleem AS, Khalid CN. Brain abscess in children. Ind J Pediatr. 2006;73:401–404. doi: 10.1007/BF02758560. [DOI] [PubMed] [Google Scholar]
- 3.Taylor TE, Fu WJ, Carr RA, et al. Differentiating the pathologies of cerebral malaria by postmortem parasite counts. Nature Med. 2004;10:143–145. doi: 10.1038/nm986. [DOI] [PubMed] [Google Scholar]
- 4.Yogev R, Bar-Meir M. Management of Brain Abscess in Children. Pediatr Infect Dis J. 2004;23:157–159. doi: 10.1097/01.inf.0000110272.67271.a2. [DOI] [PubMed] [Google Scholar]
- 5.Sáez-Llorens X. Brain Abscess in Children. Semin Pediatr Infect Dis. 2003;2:108–114. doi: 10.1053/spid.2003.127227. [DOI] [PubMed] [Google Scholar]
- 6.Brook I. Brain Abscess in children: microbiology and management. J Child Neurology. 1995;10:283–288. doi: 10.1177/088307389501000405. [DOI] [PubMed] [Google Scholar]
- 7.Bronzan RN, Taylor TE, Mwenechanya J, et al. Bacteremia in Malawian children with severe malaria: prevalence, etiology, HIV coinfection and outcome. J Infect Dis. 2007;195:895–904. doi: 10.1086/511437. [DOI] [PubMed] [Google Scholar]
- 8.Mankhambo L, Chiwaya K, Phiri A, Graham SM. Lobar pneumonia caused by nontyphoidal Salmonella in a Malawian child. Pediatr Infect Dis J. 2006;25:1190–1192. doi: 10.1097/01.inf.0000245098.82276.d6. [DOI] [PubMed] [Google Scholar]
- 9.Beare NAV, Taylor TE, Harding S, Lewallen S, Molyneux ME. Malarial retinopathy: a newly established diagnostic sign in severe malaria. Am J Trop Med Hyg. 2006;75:790–797. [PMC free article] [PubMed] [Google Scholar]