

* Departments of Neurology and Pediatrics, University of Virginia, Charlottesville, Virginia
Divisions of Emergency Medicine and Infectious Diseases, Childrens Hospital, Boston, Massachusetts
Department of Neurology, Childrens Hospital, Harvard Medical School, Boston, Massachusetts
| ABSTRACT |
|---|
|
|
|---|
Methods. To ensure that all occurrences of intracerebral abscess treated at Childrens Hospital Boston between 1981 and 2000 were identified, we searched 4 separately maintained databases for the keywords "brain" or "abscess" or the International Classification of Diseases, Ninth Revision code 324.x. This search yielded the names of 386 patients. Of these 386 patients, a solitary intracerebral abscess or multiple noncontiguous intracerebral abscesses could be confirmed in 54 patients on the basis of cranial imaging (computed tomography or magnetic resonance imaging) or autopsy reports. The complete retrospective review of the medical records of these 54 patients constitutes the basis for this study.
Results. Congenital heart disease was the most common predisposing factor during both eras. Compared with the previous era, important historical trends identified include a reduction in the number of abscesses that occurred in the settings of sinus or otitic infection (11% during 19812000 vs 26% during 19451980), an increase in number of intracranial abscesses in infants (18% vs 7%) and in the setting of immunosuppression (16% vs 1%), an increase in the number of children who were treated with antibiotics alone (22% vs 1%), a stable overall mortality rate (24% vs 27%), and the identification of Citrobacter and fungus as causes of intracranial abscess not observed during the previous era of 19451980. Citrobacter was observed only during the neonatal period. Fungi were the causative organisms predominantly in the setting of immunosuppression.
Conclusions. Intracerebral abscess in children continues to be associated with high rates of neurologic impairment and death. Because earlier detection may reduce morbidity and mortality, intracranial abscess should be considered when evaluating children with new-onset neurologic signs or symptoms, especially in children who have acute immunosuppression and disseminated fungal disease or fungemia.
Key Words: abscess brain children
Abbreviations: VP, ventriculoperitoneal CT, computed tomography MRI, magnetic resonance imaging CI, confidence interval
Intracerebral abscesses result from the invasion of infectious organisms into the parenchyma of the brain as a consequence of the spread of contiguous infection from nonneural tissue, the result of hematogenous introduction from a remote site, or direct mechanical introduction as a result of penetrating trauma or a surgical procedure. A wide range of microorganisms have been recovered from intracerebral abscesses, including most types of bacteria and certain types of fungi and parasitic organisms.13
A number of previous studies have characterized the natural history of intracerebral abscess in children.48 Predisposing factors identified in these studies include congenital heart diseases; infections of the middle ear, mastoids, paranasal sinuses, and soft tissues of the face, orbit, or scalp; penetrating skull injury; comminuted fracture of the skull; intracranial surgery, including the insertion of ventriculoperitoneal (VP) shunts; congenital lesions of the head and neck, including dermal sinuses usually located over the posterior fossa; and abnormalities of immune system function. In some cases, no predisposing factors or source of infection can be found.9
These previous studies documented that intracranial abscesses often cause significant morbidity, including epilepsy, motor or sensory dysfunction, visual field defects, and personality change.46 Mortality was high in the mid-1900s, but it subsequently declined. At Childrens Hospital Boston, mortality decreased from 36% in 19451970 to 14% in 19711980.4 It is likely that this improvement is attributable to such factors as the introduction of cranial computed tomography (CT) scanning, improvement of neurosurgical technique, and advances in antimicrobial therapy.8,10,11
In the present study, we assessed the clinical presentation and the outcome of intracerebral abscesses in children who presented to Childrens Hospital Boston between 1981 and 2000 and compare our results with the findings from a previous study from 1945 to 1980.4 We were particularly interested in 1) whether the number of episodes of intracerebral abscess associated with immunoincompetence, a factor observed in
1% of the episodes during the earlier epoch of 19451980, has increased in our era of a considerably increased prevalence of immunosuppression as a result of organ transplantation, aggressive cancer therapy, and human immunodeficiency virus infection; 2) whether there were changes in the microbiology of intracerebral abscesses; and 3) whether mortality rate at our institution had continued to decline.
| METHODS |
|---|
|
|
|---|
Because we could not ensure that all patients with intracerebral abscesses received a primary discharge diagnosis code of intracerebral abscess, search of this database was supplemented by search of 3 additional databases: 1) a radiology database of cranial CT and magnetic resonance imaging (MRI) reports for the word "abscess," 2) a neuropathology database of diagnoses for the word "abscess," and 3) a database of microbiology culture reports for the term "brain." This search strategy yielded the names of an additional 345 children whose records were then reviewed. Only 13 of these children were found to have either radiographic evidence (CT or MRI) or pathologic evidence at autopsy of 1 or more intracerebral abscesses in all stages of development.
The complete review of the medical records of the 41 children who were identified by search of the medical record database of primary discharge diagnosis and the additional 13 children who were identified by the search of the supplemental databases constitutes the basis for this study. These records were reviewed for information concerning demographics, predisposing factors, presenting symptoms, site of abscess, organism, treatment, and outcome. Among the 332 who were excluded from additional study, patients were excluded when 1) radiologic studies were performed to exclude the presence of an abscess or 2) radiologic or pathologic studies demonstrated an abscess that was extracerebral in location. This study was approved by the Childrens Hospital Boston Institutional Review Board.
| RESULTS |
|---|
|
|
|---|
Predisposing Factors
The patients ranged from 5 days to 34 years of age. The predisposing factors associated with intracerebral abscess development between 1981 and 2000 are listed in Table 1. To allow comparison with the earlier era, the predisposing factors identified between 1945 and 1980 are presented in the last column of Table 1.
|
Of the 10 infants in this study, 7 were younger than 6 weeks at presentation with abscess and 5 were younger than 1 month. Five of the 10 infants were born prematurely (<37 weeks gestational age). One of these children required the placement of a VP shunt in the perinatal period. During the 19451980 epoch, only 2 of the 7 infants were younger than 6 weeks (2 and 5 weeks of age) at presentation with abscess. The gestational age of these infants is not mentioned in the prior study.
Nine children from the 19812000 epoch are classified as immunosuppressed. Six of these children had organ transplantations (2 renal, 1 lung, 1 liver, 2 bone marrow). Two children were undergoing induction chemotherapy for acute lymphoblastic leukemia. One child had hyperimmunoglobulin M syndrome. We classified 1 of the children during the 19451980 epoch, a heart transplant recipient, as immunosuppressed. Furthermore, we considered 2 adolescent children who had previously been classified as immunocompromised as a result of cystic fibrosis to be immunocompetent.
Abscess Location
Thirty-seven children during the 19812000 era were found to have a single abscess; 18 had multiple noncontiguous abscesses. Whether single or multiple, most abscesses were located in the cerebral hemispheres (Table 2). Four of the children were noted to have abscesses involving both the cerebral hemispheres and subcortical structures. Three of these were immunosuppressed children who had fungal abscesses in association with systemic fungal infection. The fourth child, who had trisomy 21 and congenital heart disease, developed multiple abscesses in association with Staphylococcus aureus septicemia. All 4 of these children died. Because cerebral localization, as defined by head imaging, is much more refined in the current era and patients with multiple abscesses during the earlier era were probably missed, a direct comparison between the 2 eras is not possible.
|
The organisms identified from any culture (including those obtained at the time of autopsy) are listed in Table 3. Members of the Citrobacter genus were the causative organism in 3 of the 5 children younger than 1 month. One childs abscesses were the result of Citrobacter diversus. In the other 2 episodes, Citrobacter was identified as the causative organism before transfer to Childrens Hospital. The subtype was not noted in the available medical record. Citrobacter was not identified as cause of a brain abscess during the 19451980 epoch.
|
Abscesses as the result of fungal infections occurred in 9 of the children in the 19812000 epoch. Seven of these children were immunosuppressed. The eighth child was a neonate with intestinal neuronal dysplasia. The ninth child was a 4-year-old with a VP shunt. Seven of these children died. Fungi were not identified as a causative agent during the 19451980 era.
Clinical Presentation
Neurologic signs or symptoms, in isolation or combination, were noted in 52 of the 55 patients who presented with abscesses between 1981 and 2000 (Fig 1). On the basis of the history available in the medical records, headaches (n = 27; 50%) were all of new onset and were associated with 1 or more of the following: vomiting (n = 12), photophobia (n = 5), and fever (n = 16). None of the children who presented with seizures (n = 15; 27%) had a known history of febrile seizures or epilepsy. Mental status changes, ranging from lethargy to coma, occurred in 17 children. Two children presented after apparent life-threatening events. Other neurologic findings included paresthesias (n = 4), dysphagia (n = 2), dysarthria (n = 2), aphasia (n = 1), weakness or hemiparesis (n = 4), nystagmus (n = 1), developmental regression (n = 1), and increasing head circumference (n = 3).
|
The type and frequency of these various findings are similar to those of the 19451980 epoch in which headaches were present in 72%, seizures in 36%, and other neurologic symptom or deficits in 59%. However, 13% (n = 12) of the abscesses in the 19451980 epoch presented in coma as compared with 3 (5%) children in the 19812000 group. The comatose patients in the current era presented to the emergency departments of other facilities with complaints of new-onset headaches and confusion (n = 2) or new-onset seizure and right hemiparesis (n = 1). Coma developed before transfer to Childrens Hospital. One of these episodes resulted in death from herniation. The other 2 in this group eventually recovered.
Treatment
One patient was not treated with antibiotic. All other patients received antimicrobial therapy alone or in combination with a surgical procedure. The group that received antimicrobial therapy alone (n = 12) included 1) 3 neonates with multiple intracerebral abscesses and bacteremia; 2) 1 child with multiple intracerebral abscesses and a biopsied lung abscess from which S milleri was isolated; 3) 3 children with small (<2 cm in diameter), solitary, intracerebral abscesses; 4) 2 children who had small, intracerebral abscesses and also had subdural empyemas that were surgically drained; and 5) 3 children whose intracerebral abscesses were detected by cranial CT in the presence of septic shock, multiple system failure, and/or fungemia just before death. During the 19451980 era, only 1 child was treated with antibiotic therapy alone.
A surgical treatmentaspiration (n = 39) or resection (n = 3)was performed in 42 of these 55 cases. Second (n = 13), third (n = 6), and fourth (n = 1) procedures were required in the 19812000 group as a result of new neurologic symptoms or continued abscess growth despite antibiotic therapy deemed appropriate on the basis of organism susceptibility profiles. A similar fraction of patients required multiple surgical procedures in the 19451980 group.
Surgical complications included hemorrhage (n = 4), stroke (n = 2), and cerebrospinal fluid leak (n = 2). Three of the 4 children who had surgical intervention complicated by hemorrhage died. There was no apparent relationship between types of organism and requirement for 1 or more surgical procedures or need for resection in the 19812000 group.
Outcome
Thirteen of the 19812000 patients died. The final outcome of 1 child who was discharged to hospice care in coma is not known. Sixteen children did not return to Childrens Hospital Boston for outpatient care or subsequent admissions.
Morbidity
Outcome information is available for 24 patients. Seven of the children returned to their premorbid baseline function. Of the remaining 15 patients, new developmental delays or learning disorders were noted in 10. Six children developed epilepsy, 4 developed new motor deficits, 1 developed a persistent hemianopsia, and 3 developed hydrocephalus requiring the placement of a VP shunt.
Mortality
As noted above, 13 of the 55 abscess episodes (24%; 95% CI: 13%37%) resulted in death before hospital discharge. The ages of these patients ranged from 1 month to 34 years. Specifics concerning these patients are presented in Table 4. All received antimicrobial therapy. Five (patients 1, 2, 6, 9, and 10) received antimicrobial therapy without surgical intervention. In the others, a single aspiration was performed. In patients 5, 11, and 12, aspiration of the abscess was complicated by hemorrhage.
|
| DISCUSSION |
|---|
|
|
|---|
By comparing our current experience with this earlier published data, several observations can be made. First, congenital heart disease remained the most common predisposing factor associated with the development of intracerebral abscess at our institution. Second, fewer abscesses in the 19812000 epoch occurred in the setting of sinus and otitic infections. Third, more episodes during the modern era arose in the setting of immunosuppression. The immunosuppression was usually acute, and the intracranial abscess was usually a severe consequence of disseminated fungal disease. Most of these patients (67% [6 of 9]) died. Fungal abscesses were not identified during the previous era. Fourth, more neonates with intracerebral abscesses were identified between 1981 and 2000. In 3 of these neonates, the causative organism was Citrobacter, an agent not identified during the previous era. Fifth, presentation in coma was less common in the modern era. Sixth, more children during the modern era were treated with antimicrobials alone.
We suspect that the decline in sinus and otitic infections resulting in intracerebral abscesses is the direct result of improvement in diagnosis and antimicrobials as well as the increased use and efficacy of antibiotics for the treatment of sinus and otitic infections between 1981 and 2000.16 With increasing trends to limit the prescription of antibiotics for sinus and otitic infections,17 it is possible that we will see a resurgence of these infections as a predisposing factor for intracerebral abscess.
Not surprising, the number of children with immunosuppression as the predisposing factor has increased compared with the previous study.3,18,19 Bone marrow transplantation and solid organ transplantation were virtually nonexistent during the majority of the years of the previous study. Of the 9 children who were identified as immunosuppressed in the 19812000 group, 8 were acutely immunosuppressed and 7 had systemic fungal infections.
Availability of improved neuroimaging techniques has likely led to earlier and more accurate diagnosis. In the earlier era, 6 patients died before diagnosis or commencement of treatment. This was not true of patients in the 19812000 group. It also likely has increased the number of children younger than 6 months diagnosed with intracerebral abscess.
The treatment strategies of the earlier era are comparable to those used in the 19812000 group except for an increased tendency in the current era for treatment with antimicrobials alone. This difference may reflect earlier detection and a trend toward treating those who are neurologically intact and the abscess or abscesses are small (<2 cm) with antimicrobials alone.2022
Analysis of 19451980 cases showed reduction in mortality rates after 1970. This improved outlook was attributed to the introduction of noninvasive cranial imaging techniques.8,10 We did not find any additional decline in the mortality rate at our institution between 1981 and 2000 compared with the interval after 1970. However, the most common cause of death changed from intracranial causes during the earlier era to multisystem failure between 1981 and 2000.
| CONCLUSION |
|---|
|
|
|---|
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
Reprint requests to (H.P.G.) Department of Neurology, Box 800394, University of Virginia, Health Science Center, Charlottesville, VA 22908. E-mail: hpg9v{at}virginia.edu
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
M. Epelman, A. Daneman, S. I. Blaser, C. Ortiz-Neira, O. Konen, J. Jarrin, and O. M. Navarro Differential Diagnosis of Intracranial Cystic Lesions at Head US: Correlation with CT and MR Imaging RadioGraphics, January 1, 2006; 26(1): 173 - 196. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. G. Millichap Prevalence, Symptoms, and Prognosis of Intracerebral Abscess AAP Grand Rounds, August 1, 2004; 12(2): 15 - 15. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||