Ovid - Current Treatment of Brain Abscess in Patients With Congenital Cyanotic Heart Disease.

Ovid: Current Treatment of Brain Abscess in Patients with C...
http://ovidsp.tx.ovid.com.ezp.lib.unimelb.edu.au/sp-3.18.0b/o...
Neurosurgery
Issue: Volume 41(6), December 1997, pp 1270-1279
Copyright: Copyright by the Congress of Neurological Surgeons
Publication Type: [Clinical Studies]
ISSN: 0148-396X
Accession: 00006123-199712000-00006
Keywords: Brain abscess, Cyanotic heart disease, Infection, Mortality
[Clinical Studies]
Current Treatment of Brain Abscess in Patients with Congenital Cyanotic Heart Disease
Takeshita, Mikihiko MD; Kagawa, Mizuo MD; Yato, Seiji MD; Izawa, Masahiro MD; Onda, Hideaki MD; Takakura, Kintomo MD; Momma, Kazuo MD
Author Information
Department of Neurosurgery, Neurological Institute (MT, MK, SY, MI, HO, KT), and Department of Pediatric Cardiology (KM), The Heart Institute of Japan, Tokyo
Women's Medical College, Tokyo, Japan
Received, August 13, 1996.
Accepted, July 16, 1997.
Reprint requests: Mikihiko Takeshita, M.D., Department of Neurosurgery, Neurological Institute, Tokyo Women's Medical College, Kawada-cho 8-1, Shinjuku-ku,
Tokyo 162, Japan.
Abstract
OBJECTIVE: The goal of this study was to define clearly the role of management in patients with cyanotic heart disease and brain abscesses by evaluating
retrospectively the factors influencing poor outcome in these patients.
METHODS: This study included 62 patients with cyanotic heart disease and brain abscesses diagnosed in the computed tomography era. Basic
characteristic parameters (number, size, location, computed tomographic classification and organism type of abscess, convulsion, type of cyanotic heart
disease, age distribution, immunocompromised status, pretreatment neurological state, and intraventricular rupture of brain abscess [IVROBA]) and
therapeutic parameters (type of antibiotics and duration of administration, steroid medication and therapeutic modalities, aspiration with or without
cerebrospinal fluid drainage, total extirpation after aspiration, or primary extirpation and medical treatment) were evaluated as independent predictors of
poor outcome (totally disabled state or death) by using univariate and multivariate logistic regression analysis. We also statistically estimated the possible
causes of IVROBA and the multiplicity of brain abscess.
RESULTS: Although there were no statistically significant correlations between patients with good and poor outcomes in regard to other basic characteristic
and therapeutic parameters, patients with poor outcomes were older(P < 0.02), more frequently had IVROBA (P < 0.005), and had a higher frequency of
neurological deterioration (P< 0.01) than those with good outcomes. Multiple logistic regression analysis predicted that poor outcome increased the relative
risk of IVROBA by a factor of 18.9 (odds rate, 18.9; 95% confidence interval, 1.7-211.6;P < 0.02). More patients with multiple abscesses had positive
immunocompromised states than those with single abscesses (P < 0.01). Deep-located abscesses also more frequently had IVROBA (P< 0.005) and
abscesses located in the parieto-occipital region ruptured into the occipital horn of the lateral ventricle in a short period(P < 0.02).
CONCLUSIONS: Our findings suggest that IVROBA strongly influences poor outcome in patients with cyanotic heart disease. The key to decreasing poor
outcomes may be the prevention and management of IVROBA. To reduce operative and anesthetic risk in these patients, abscesses should be managed by
less invasive aspiration methods guided by computed tomography. Abscesses larger than 2 cm in diameter, in deep-located or parieto-occipital regions,
should be aspirated immediately and repeatedly, mainly using computed tomography-guided methods to decrease intracranial pressure and avoid IVROBA.
IVROBA should be aggressively treated by aspiration methods for the abscess coupled with the appropriate intravenous and intrathecal administration of
antibiotics while evaluating intracranial pressure pathophysiology.
Because computed tomography (CT) facilitates early detection, accurate characterization, and evaluation of brain abscesses and intracranial pressure
pathophysiology, the combination of surgical and medical management has dramatically reduced the mortality rates from 40 to 60% (3,6,23,25) in the pre-CT
era to the current rates of 0 to 10% (17,23,28,35,37). The application of stereotactic aspiration using CT for deep-seated or multiple abscesses in recent years
has also improved mortality and morbidity rates (8,13,17,31,33).
Patients with brain abscess and congenital cyanotic heart disease, however, possess not only cardiopulmonary risk but also a wide variety of coagulation
defects. Together, these factors increase the risks of anesthesia and surgery (9,20,27,32). Radford et al. (22) and Radford and Thong (21) also demonstrated
that children with congenital heart disease have variable degrees of immunodeficient states.
Cyanotic brain abscesses caused by hematogenous dissemination from a primary silent infected focus are often deep-seated and located at the junction of
white and gray matter. The capsule is generally thin on the ventricle side compared with the portion near the cortical surface (30). Consequently, the growth of
the abscess may cause a rupture into the ventricles accompanied with mass eect. Even today, intraventricular rupture of brain abscess (IVROBA) continues to
be associated with a high mortality rate (38).
Despite improvements in imaging techniques and in surgical and medical management with newly developed antibiotics, the mortality rate for cyanotic brain
abscess (20) is 13.3%, even in the CT era. Because these patients have unstable hemodynamic conditions and coagulation disorders and the brain abscesses
may easily rupture into the ventricle, there is much controversy over the best way of treating these patients. For example, the number, location, and size of the
brain abscesses that should be surgically excised or aspirated (24,26) and the exact duration of antibiotic therapy and the relative role of corticosteroids remain
unclear. The abscess diameter on the necessity of repeated aspiration and the need for placement of the ventricular drainage when the abscess has ruptured
into the ventricle are also a matter of argument (38). There is little information available to guide the management of patients with cyanotic brain abscess, and
the management of these patients has been subjective to greater variability.
1 of 12
1/24/16, 10:44 AM
We reported in 1983 on 62 patients with brain abscess and congenital cyanotic heart disease in both the pre-CT and CT eras (12). We predicted in that article
that the number of patients cured by aspiration with intravenous antibiotics administration would increase after the introduction of CT. We have used the
therapeutic modality of repeating aspiration of abscesses concomitantly combined with the administration of intrathecal and intravenous antibiotics after
IVROBA in the CT era. To establish the best management for cyanotic brain abscess, we statistically determined the basic characteristics and therapeutic
factors influencing outcome in these patients in the CT era and the optimal management of these problematic lesions and demonstrated the basic
characteristics of total cyanotic brain abscess in both the pre-CT and CT eras.
PATIENTS AND METHODS

From May 1969 to April 1995, 149 patients with brain abscesses were treated at the Department of Neurosurgery, Neurological Institute, Tokyo Women's
Medical College. Of these 149 patients, 103 (69.1%) had congenital cyanotic heart disease. These patients included five with tetralogy of Fallot, who
developed new abscesses in another region after resolution of the initial abscesses after 3 to 5 years. All patients in this series had abscesses before radical
operation for the heart disease. Thirty-nine patients received palliative shunts (Blalock-Taussing anastomosis in 33 patients, Waters anastomosis in 3, Glenn's
operation in 1, and atrioseptostomy in 2) for the heart disease at ages of 0.3 to 11 years. The brain abscesses developed between 1 month and 29 years
(mean, 8.9 yr) after a palliative shunt.
Between 1969 and 1975, CT was not yet routinely used. Stereotactic aspiration procedures guided by CT for high-risk locations or multiple abscesses as
variable treatment options became available from 1984 onward.
Basic characteristics of total cyanotic brain abscess in the pre-CT and CT eras (Table 1)
TABLE 1. Initial Symptoms and Clinical Characteristics in 103 Patients with Cyanotic Brain Abscessesa
2 of 12
1/24/16, 10:44 AM
Of the 103 patients, 39 were female and 64 were male patients. Their ages ranged from 1 to 56 years, and 58 patients were 10 years old or younger. The most
common form of congenital cyanotic heart disease was tetralogy of Fallot in 57 patients, and the next most common were complete transposition of the great
arteries in 12 and double outlet of the right ventricle in 10. The most common clinical features were symptoms and signs of raised intracranial pressure,
together with pyrexia. The main localizing sign was hemiparesis, and epileptic seizure occurred in 48 patients (46.6%). Sixty-seven patients developed
disturbed consciousness, but semicoma occurred in only 6.8%.
Most abscesses were located supratentorially, and 19.4% were multiple abscesses. Each of 83 patients had a solitary abscess, including 12 multilocular
lesions. Each of 14 patients had two separate lesions, each of 5 patients had three discrete abscesses, and 1 patient harbored five discrete abscesses.
Information concerning the bacteriology of the cyanotic brain abscess was not fully available during the first 3-year investigational period. Thirty-nine of 76 pus
cultures (51.3%) showed microorganisms, with single organisms in 30 patients and mixed infection in 9 patients. During the period from 1971 to 1995,
anaerobic organisms (in pure or mixed culture) were isolated in 17 patients (22.4%). The most common aerobic bacteria wereStaphylococcus species,
Streptococcus species, and Gram-positive cocci, whereas the most common anaerobic bacteria werePeptostreptococcus. Blood cultures in 39 patients were
mostly sterile (89.8%), and microorganisms of the brain abscesses matched those of the blood cultures in only two patients.
Basic characteristics of cyanotic brain abscess in the CT era (Table 2)
TABLE 2. Brain Abscesses in Patients with Congenital Cyanotic Heart Disease (Computed Tomography Era)
3 of 12
1/24/16, 10:44 AM
We statistically determined the basic characteristics and therapeutic factors influencing outcome in 62 patients with cyanotic brain abscesses in the CT era. Of
the 62 patients, 24 were female and 38 were male patients. Their ages ranged from 1 to 56 years. The staging of brain abscess by Britt and Enzmann (4), which
correlates histological features of brain abscess, demonstrated late cerebritis in 31 patients and early or late capsule in 31 patients. Pretreatment neurological
state was classified as alert state and/or mild deficit, lethargic state and moderate deficit, or obtunded state and marked deficit.
Treatment and complications

[beta]-Lactam antibiotics were administered intravenously and immediately to all patients to avoid IVROBA. Determining the appropriate bacteriocidal
antibiotics, with the spectrum narrowed, depends on the sensitivity of the microbial organism, which is generally identified within 48 to 72 hours. The duration
of antibiotic therapy was not influenced by the number of abscesses or whether patients underwent surgical or medical therapy. In patients who did not
undergo surgery and in whom the microorganism was unidentified, antibiotic therapy was based on results from blood culture or drugs were chosen
empirically. Only seven patients with severely disturbed consciousness received corticosteroids to reduce brain edema, with administration lasting less than 7
days.
We selected conservative treatment using antibiotics for patients with abscesses smaller than 2 cm in diameter. These patients underwent surgery if 1 week of
therapy failed to demonstrate clinical or radiographic improvement. Abscesses larger than 2 cm in diameter and causing significant mass eect were treated
surgically, mainly using aspiration methods. When we aspirated individual abscesses, 50 to 80% of their volumes were removed (Fig. 1). Repeated aspiration
was required when abscesses did not decrease to smaller than 2 cm in diameter after 2 weeks of antibiotic therapy. Two or three aspiration procedures were
generally sucient.
FIGURE 1. A, contrast-enhanced computed tomographic scans of the head showing a ring-enhancing multilocular brain abscess at the right frontal lobe in a 23-year-old
man with the single ventricle. B, contrast-enhanced computed tomographic scans of this patient showing residual brain abscess after stereotactic aspiration.
When a cyanotic brain abscess ruptured into the ventricle, we first chose medical treatment for the abscess, evaluating the abscess and intracranial pressure
pathophysiology by performing weekly CT. We secondarily treated the abscesses that were larger than 2 cm in diameter accompanied with dilated ventricles
by aspiration of the abscesses and ventricular drainage coupled with the administration of appropriate intravenous and intrathecal (mainly gentamycin)
antibiotics (Fig. 2). Although the administration of intrathecal antibiotics with intermittent or continuous spinal drainage was limited to patients with abscesses
smaller than 1.5 cm in diameter with the capsule separated from the ventricle, we had no cases of subsequent IVROBA with this treatment. We do not
stereotactically aspirate the small cyanotic brain abscess (<1.5 cm) for purposes of diagnosis, because these patients are at high surgical and anesthetic risk.
FIGURE 2. A, preoperative contrast-enhanced computed tomographic scan in a patient with right deep parieto-occipital lobe and IVROBA. B, after the aspiration of the
abscess and placement of ventricular drainage stereotactically, the right deep parieto-occipital abscess decreased in size.
Fourteen patients were treated conservatively, and four with IVROBA were supplemented by the administration of intrathecal antibiotics with intermittent spinal
drainage. Thirty-nine of 48 surgical patients underwent repeated aspiration, including 10 patients with continuous ventricular drainage, 2 with continuous or
intermittent spinal drainage, and 2 with external decompression. Five of the patients with IVROBA underwent repeated aspiration only. In the early period after
the introduction of CT, patients underwent primary excision or excision after aspiration when the abscess did not decrease in size. For statistical analysis,
therapeutic modalities were classified as aspiration with or without drainage (including ventricular drainage only), extirpation after aspiration, or primary
extirpation and conservative treatment.
4 of 12
1/24/16, 10:44 AM
Although complications from broad spectrum antibiotics included sepsis caused by fungal superinfection in one patient with depressed immune response,
there were no patients with pseudomembraneous colitis. Although three patients had intracapsular hematomas after aspiration and were treated
conservatively, none died as a result of the hematomas. One patient with IVROBA, however, had intracapsular and intraventricular hematomas after aspiration
and ventricular drainage and died as a result of uncontrollable ventriculitis.
Statistical analysis
Statistical analysis was performed using the SPSS package (26). In univariate analysis, the [chi]2 or Fisher's exact test was used to compare proportions.
Multiple logistic regression models were used to test and quantify relations independent of possible theoretical risk factors influencing poor outcome (totally
disabled state or death). We then performed a multivariate logistic regression analysis to determine whether any of the variables determined to be related in the
univariate analyses (P < 0.1) would be independently related to exclusion from this study. Only factors with a potential relationship to dependent variates were
included in the model. The odds rates(OR) with 95% confidence intervals (95% CIs), which were estimated using the standard errors, were calculated by
logistic regression before and after adjustment for potential confounding variables. Significance was judged at the P < 0.05 level.
RESULTS
Basic characteristics and therapeutic factors influencing outcome of brain abscess
The clinical features and treatment in patients with cyanotic brain abscesses with good versus poor outcomes in the CT era are compared in Table 2. Patients
with poor outcomes were older(P < 0.02), more frequently had IVROBA (P < 0.005), and had a higher frequency of neurological deterioration (P< 0.01) than
those with good outcomes. There was no statistically significant correlation between patients with good and poor outcomes with respect to the number, size,
computed tomographic classification, organism type of abscess, convulsion, type of cyanotic heart disease, or immunocompromised status. Patients with
poor outcomes had deep-located abscesses more often than patients with good outcomes, but this dierence was not significant.
Therapeutic factors such as the type of antibiotics and duration of administration, steroid medication, and therapeutic modalities were not statistically
significant as collected in patients with poor or good outcomes. Only factors with a potential relationship to dependent variates were included in the
multivariate logistic regression model (age distribution, abscess in the noneloquent versus eloquent locations, IVROBA, and pretreatment neurological grade).
IVROBA occurred more frequently in patients with poor outcomes. The logistic regression analysis determined that poor outcome increased the relative risk of
IVROBA by a factor of 18.9 (OR, 18.9; 95% CI, 1.7-211.6, P < 0.02).
Patients with poor outcomes were statistically significantly older than patients with good outcomes. Patients older than 21 years, however, did not have a
higher relative risk than patients aged 11 to 20 years (OR, 1.8; 95% CI, 0.39-8.8). Patients aged 11 to 20 years were also less likely to have poor outcomes than
patients younger than 10 years (OR, 0.2; 95% CI, 0.06-1.1).
Although patients with poor outcomes more frequently had abscesses in the eloquent area than those with good outcomes, logistic regression analysis
demonstrated that the former did not have a relatively higher risk of such abscesses. Patients with poor outcomes had a high frequency of neurological
deterioration. Obtunded state and marked neurological deficit did not predict poor outcome more than lethargic state and moderate neurological deficit (OR,
0.8; 95% CI, 0.2-2.8). Patients with alert state and/or mild neurological deficits also did not have poorer outcomes than patients with lethargic status and
moderate neurological deficits (OR, 0.3; 95% CI, 0.06-2.0).
Number of abscesses (single versus multiple abscess) and clinical features

Table 3 demonstrates the relationship between single and multiple abscesses in these patients. Although more patients with multiple abscesses were
immunocompromised than patients with single abscesses(P < 0.01) and more frequently had IVROBA (P < 0.005), there was no statistically significant
correlation between patients with multiple abscesses and single abscesses with respect to other clinical features and outcome.
5 of 12
1/24/16, 10:44 AM
TABLE 3. Relationship between Single and Multiple Abscesses in Patients with Congenital Cyanotic Heart Disease (Computed Tomography Era)
IVROBA in the CT era

Table 4 shows the clinical features in cyanotic brain abscesses patients with and without IVROBA. IVROBA occurred in 21 patients(33.9%), resulting in four
deaths and one poor outcome, whereas the mortality rate of patients without IVROBA was 7.3%. There were no significant dierences in the size, number,
computed tomographic classification of abscess, or type of cyanotic heart disease between patients with and without IVROBA, although deep-located
abscesses more frequently ruptured intraventricularly (P < 0.005).
6 of 12
1/24/16, 10:44 AM
TABLE 4. Clinical Features of Intraventricular Rupture of Cyanotic Brain Abscess in the Computed Tomography Eraa
Even in the CT era, 17 patients had IVROBA before being referred to the Department of Neurosurgery of the Tokyo Women's Medical College and 4 patients
had IVROBA during their stay in our clinic. Relationships between the ventricular site at which an abscess ruptured and the span from initial symptom to
IVROBA are compared in Table 5. Abscesses located in the parieto-occipital region may rupture into the occipital horn of the lateral ventricle in a short period
(P < 0.02).
TABLE 5. Relationship between the Portion of Intraventricular Rupture of Brain Abscess and a Span from the Beginning of Symptomatologya
Convulsion and anticonvulsants

Anticonvulsants were administered to all patients diagnosed with brain abscess. All patients received anticonvulsants for periods ranging from 6 months to 10
years. The incidence of seizure was 44% at presentation (versus 51% in the pre-CT era), 11% during treatment (versus 2.4% in the pre-CT era), and 37% in
the CT era (versus 30% in the pre-CT era). There were no significant dierences between the incidence of seizure in the pre-CT era and in the CT era.
DISCUSSION
The incidence of brain abscess in the population with congenital heart disease varies from 5 to 18.7% according to previous reports (9), and high mortality
rates for cyanotic brain abscess have been reported in both the pre-CT and CT eras (10,12,17,36). The introduction of CT and new antibiotics has dramatically
reduced mortality rates for total purulent brain abscesses from 40 to 60% (3,6,23,25) in the pre-CT era to the current rates of 0 to 10% (17,23,28,35,37). A
recent report on cyanotic brain abscess, however, states that the mortality rate is 13.3%, even in the CT era (20).
Our series of 62 patients with cyanotic brain abscesses in the CT era represents the largest series of such patients reported to date in a country where surgery
for cyanotic heart disease has been conducted aggressively. On the basis of the results of this series, there was no statistically significant correlation between
patients with good and poor (dead or totally disabled) outcomes determined by univariate analysis regarding clinical features and therapeutic modalities except
for age distribution, pretreatment neurological state, and IVROBA. Multivariate logistic regression analysis determined IVROBA to be a high risk factor for poor
outcome in patients with cyanotic brain abscesses.
We think these results may be largely attributable to the accuracy with which CT detected and localized cyanotic brain abscesses, coupled with prompt
aspiration of any and all abscesses larger than 2 cm in diameter, which causes significant mass eect, even in the late cerebritis stage. Furthermore, we have
aggressively treated brain abscess accompanied with IVROBA by aspiration, coupled with the administration of appropriate intravenous and intrathecal
antibiotics.
7 of 12
1/24/16, 10:44 AM
Patients with cyanotic brain abscesses, which may develop in childhood and adolescence, are at high cardiopulmonary risk (9,20,27,32) and possess a wide
variety of coagulation defects and variable degree of immunodeficient states (21,22). For these patients, less invasive surgical procedures such as aspiration
should be chosen to reduce operative and anesthetic risks.
Etiology of cyanotic brain abscess

The etiology of cyanotic brain abscess presents several problems. Intracardiac right-to-left shunt bypass, by which blood is not filtered through pulmonary
circulation where bacteria are intercepted by phagocytosis, may allow direct entry to cerebral circulation (9,12,18). In these patients, the brain may also have
predisposing conditions such as minute low-perfusion areas because of severe hypoxemia and metabolic acidosis and increased blood viscosity resulting from
compensatory polycythemia (9,18), which, together with a consequent low-perfusion area, prepares a focus receptive to infection. Shunted blood containing
microorganisms may be seeded in such lesions, forming a cerebral abscess. This hypothesis implies that infarcts other than at the site of the abscess should
be observed or multiple metastatic brain abscesses should occur.
Fischbein et al. (9) and we, in a previous report (27), retrospectively evaluated risk factors for cyanotic brain abscess in patients with this condition compared
with a control group with cyanotic heart disease without brain abscess. The mean arterial oxygen saturation at the ascending aorta was significantly lower than
in the patients in the control group. Accordingly, increasing the mean arterial oxygen saturation at the ascending aorta by the more eective palliative shunt
operation and prevention of cerebral emboli exaggerated by dehydration and mild infection in patients with this disease may reduce the risk of abscess
formation until radical operation for the heart anomaly (24).
Antibiotics
During the last 20 years, antibiotics and techniques for culturing anaerobic bacteria (1,7) have much improved. Sterile cultures were reported in 16 to 68% of
brain abscesses with congenital heart disease in previous reports (9,10,20,23,36) and our series (48.7%). To avoid IVROBA, which may strongly influence poor
outcome in patients with cyanotic brain abscess, [beta]-lactam antibiotics should be administered intravenously and immediately, although broad spectrum
antibiotic therapy and the use of antibiotics before surgical intervention may increase the percentage of sterile abscesses. Determining the appropriate
bactericidal antibiotics, with the spectrum narrowed, should depend on the sensitivity of the microbial organism, which is generally identified within 48 to 72
hours. In patients who did not undergo surgery and in whom the microorganism was not identified, antibiotic therapy was based on either results of blood
culture or the empirical selection of antibiotics. Antibiotics may be continued for a minimum of 6 weeks (16) and in immunocompromised patients often for 12
weeks or longer. Although complications from broad spectrum antibiotics included sepsis caused by fungal superinfection in one patient who had a depressed
immune response, there were no patients with pseudomembraneous colitis in our series.
Management of cyanotic brain abscess

In our series, the mortality rate of patients with cyanotic brain abscesses decreased significantly, to 12.9% in the CT era, but has remained consistently high in
recent years compared with purulent brain abscesses without cyanotic heart disease (17,23,28,35,37). CT is indispensable for evaluating the development of
lesions in patients for every form of therapy. For patients with abscesses smaller than 2 cm in diameter and who are in clinically stable condition, intravenous
[beta]-lactam antibiotics are recommended. Patients with abscesses, however, should undergo aspiration if 1 week of medical treatment fails to result in
clinical or radiographic improvement.
For patients with abscesses (single or multiple) larger than 2 cm in diameter or who are in a deteriorated condition, a combination of aspiration and
intravenously administered appropriate antibiotics are recommended. Particularly, abscesses located in deep and parieto-occipital regions should be aspirated
immediately and repeatedly, mainly using CT-guided methods, even in the late stage of cerebritis, in which abscesses grow rapidly with maximal edema, to
decrease intracranial pressure and avoid IVROBA. The mortality rate observed in patients with cyanotic brain abscesses without IVROBA, all of whom were
treated in the CT era, was 7.3%. This mortality rate is similar to the rates reported for solitary or multiple bacterial abscesses of other origins (8,14,17,34,35).
There has been little disagreement with our recommendation that cyanotic brain abscesses without IVROBA be managed in the same way as other purulent
abscesses (15,16,19,24).
Management of IVROBA
Cyanotic brain abscesses are frequently multiple and also occur at the junction of white and gray matter. Capsule formation is more complete on the cortical
side than the ventricular side (30), and abscesses have a propensity to rupture medially into the ventricles rather than laterally into the subarachnoid space (29),
with mass eect in the early capsular stage (4). Delay of surgical intervention is often quoted as a contributing factor for IVROBA (5). IVROBA, which is not
universally fatal but does worsen the prognosis, is often clinically demonstrated by catastrophic deterioration in the clinical status of patients with coma.
Zeidman et al. (38) collected the literature from 1950 to 1993, both for series of reported brain abscesses and for case reports with IVROBA. They claimed that
the mortality rate associated with IVROBA remained constantly high, approximately 80%, throughout this entire period. Prusty (20) reported 60 cases of
cyanotic brain abscess in the CT era (mortality rate, 13.3%), but only one patient had IVROBA, which resulted in death.
In our series, IVROBA occurred in 21 of 62 patients (33.9%), even in the CT era, and resulted in five deaths and one poor outcome (mortality rate, 23.8%). Our
findings suggest that IVROBA is a high-risk factor for poor outcome in patients with cyanotic brain abscesses in multivariate logistic analysis. Deep-located
abscesses also more frequently ruptured intraventricularly, and the abscesses located in the parieto-occipital region ruptured into the occipital horn of the
lateral ventricle in a short period.
The prevention and management of IVROBA, which is confirmed by contrast-enhanced CT or magnetic resonance imaging, may play a key role in decreasing
the incidence of poor outcome. Although previous reports concerning IVROBA of other purulent abscesses have recommended a combination of intrathecal
and intravenous antibiotic treatment, management of abscesses remains controversial. Yang and Zhao (37) request emergent craniotomy with rapid evacuation
of abscess material after IVROBA. Other reports (2,11) recommend emergent evacuation with lavage of the ventricles and ventriculostomy for intraventricular
antibiotic administration.
To reduce operative and anesthetic risks in patients with cyanotic heart disease, the first choice should be medical treatment for brain abscesses with IVROBA,
evaluating the abscesses and intracranial pressure pathophysiology by performing weekly CT. Secondary management of abscesses larger than 2 cm in
diameter and accompanied with dilated ventricles by aspiration of the abscess and ventricular drainage coupled with administration of appropriate intravenous
and intrathecal antibiotics is recommended.
CONCLUSION
8 of 12
1/24/16, 10:44 AM
Because IVROBA is strongly correlated with poor outcome in patients with cyanotic brain abscesses, the key to decreasing mortality may be prevention and
management of IVROBA. Using mainly CT-guided methods, we repeatedly aspirated abscesses larger than 2 cm in diameter, even in late cerebritis, because of
decreasing intracranial pressure and avoided IVROBA. After IVROBA, we used aggressive aspiration coupled with administration of appropriate intravenous
and intrathecal antibiotics, evaluating intracranial pressure pathophysiology by performing weekly CT. We hope that the mortality rate associated with cyanotic
brain abscess will decrease in the near future, as with other purulent brain abscesses, through the adoption of sophisticated modes of therapy.
REFERENCES
1. Alderson D, Strong AJ, Ingham HR, Path FRC, Selkon JB: Fifteen-year review of the mortality of brain abscess.Neurosurgery 8:1-6, 1981. Ovid Full Text
Bibliographic Links [Context Link]
2. Black PMcL, Levine BW, Picard EH, Nirmel K: Asymmetrical hydrocephalus following ventriculitis from rupture of a thalamic abscess.Surg Neurol
19:524-527, 1983. [Context Link]
3. Brewer NS, MacCarty CS, Wellman WE: Brain abscess: A review of recent experience. Ann Intern Med 82:571-576, 1975. Bibliographic Links [Context
Link]
4. Britt RH, Enzmann DR: Clinical stages of human brain abscesses on serial CT scans after contrast infusion: Computerized tomographic, neuropathological,
and clinical correlations. J Neurosurg 59:972-989, 1983. Bibliographic Links [Context Link]
5. Britt RH, Enzmann DR, Yeager AS: Neuropathological and computerized tomographic findings in experimental brain abscess. J Neurosurg 55:571-576,
1981. [Context Link]
6. Carey ME, Chou SN, French LA: Experience with brain abscesses. J Neurosurg 36:1-9, 1972. Bibliographic Links [Context Link]
7. de Louvois J: Bacteriological examination of pus from abscesses of the central nervous system. J Clin Pathol 33:66-71, 1980. Bibliographic Links
[Context Link]
8. Dyste GN, Hitchon PW, Menezes AH, VanGilder JC, Greene GH: Stereotaxic surgery in the treatment of multiple brain abscesses.J Neurosurg 69:188-194,
1988. Bibliographic Links [Context Link]
9. Fischbein CA, Rosenthal A, Fischer EG, Nadas AS, Welch K: Risk factors of brain abscess in patients with congenital heart disease.Am J Cardiol 34:97-102,
10. Ghosh S, Chandy MJ, Abraham J: Brain abscess and congenital heart disease. J Indian Med Assoc 88:312-314, 1990. Bibliographic Links [Context
Link]
11. Juneau P, Black PMcL: Intra-axial cerebral infectious process, in Apuzzo MLJ (ed): Brain Surgery: Complication Avoidance and Management. New York,
Churchill Livingstone, 1993, pp 1411-1418. [Context Link]
12. Kagawa M, Takeshita M, Yato S, Kitamura K: Brain abscess in congenital cyanotic heart disease. J Neurosurg 58:913-917, 1983. Bibliographic Links
[Context Link]
13. Lunsford LD, Nelson PB: Stereotactic aspiration of a brain abscess using a "therapeutic" CT scanner: A case report. Acta Neurochir (Wien) 62:25-29,
14. Mamelak AN, Mampalam TJ, Obana WG, Rosenblum ML: Improved management of multiple brain abscess: A combined surgical and medical approach.
Neurosurgery 36:76-86, 1995. Ovid Full Text Bibliographic Links [Context Link]
15. Mamelak AN, Mampalam T, Rosenblum ML: Brain abscesses: Diagnosis and management. Hosp Med 28:22-38, 1992. [Context Link]
16. Mamelak AN, Obana WG, Flaherty JF, Pharm D, Rosenblum ML: Nocardial brain abscess: Treatment strategies and factors influencing outcome.
Neurosurgery 35:622-631, 1994. Ovid Full Text Bibliographic Links [Context Link]
17. Mampalam TJ, Rosenblum ML: Trends in the management of bacterial brain abscesses: A review of 102 cases over 17 years.Neurosurgery 23:451-458,
1988. Ovid Full Text Bibliographic Links [Context Link]
18. Matson DD, Salam M: Brain abscess in congenital heart disease. Pediatrics 27:772-789, 1961. Bibliographic Links [Context Link]
19. Obana WG, Rosenblum ML: Nonoperative management of neurosurgical infection. Neurosurg Clin N Am 3:359-373, 1992. Bibliographic Links [Context
Link]
20. Prusty GK: Brain abscesses in cyanotic heart disease.Indian J Pediatr 60:43-51, 1993. Bibliographic Links [Context Link]
21. Radford DJ, Thong YH: The association between immunodeficiency and congenital heart disease. Pediatr Cardiol 9:103-108, 1988. Bibliographic Links
[Context Link]
22. Radford DJ, Lachman R, Thong YH: The immunocompetence of children with congenital heart disease. Int Arch Allergy Appl Immunol 81:331-336, 1986.
23. Rosenblum ML, Ho JT, Norman D, Weinstein PR, Pitts L: Decreased mortality from brain abscesses since advent of computerized tomography. J
Neurosurg 49:658-668, 1978. [Context Link]
9 of 12
1/24/16, 10:44 AM
24. Rosenblum ML, Mampalam T, Pons VG: Controversies in the management of brain abscess. Clin Neurosurg 33:603-632, 1986. Bibliographic Links
[Context Link]
25. Samson DS, Clark K: A current review of brain abscess.Am J Med 54:201-210, 1973. Bibliographic Links [Context Link]
26. SPSS: Statistical Package for Social Science. Chicago, SPSS Inc., Macintosh version, Release 6.0, 1994. [Context Link]
27. Takeshita M, Kagawa M, Yonetani H, Izawa M, Yato S, Nakanishi T, Monma K: Risk factors for brain abscess in patients with congenital cyanotic heart
disease. Neurol Med Chir (Tokyo) 32: 667-670, 1992. Full Text Bibliographic Links [Context Link]
28. Tekkok IH, Erbengi A: Management of brain abscess in children: Review of 130 cases over a period of 21 years. Childs Nerv Syst 8:411-416, 1992.
29. Tunkel AR, Scheld WM: Bacterial infection in adults, in Asbury AK, McKhann GM, McDonald WI (eds): Disease of the Nervous System: Clinical
Neurobiology. Philadelphia, W.B. Saunders Co., 1992, pp 1340-1349. [Context Link]
30. Waggener JD: The pathophysiology of bacterial meningitis and cerebral abscesses: An anatomical interpretation. Adv Neurol 6:1-17, 1974. Bibliographic
Links [Context Link]
31. Walsh PR, Larson SJ, Rytel MW: Stereotactic aspiration of deep cerebral abscesses after CT-directed labeling. Appl Neurophysiol 443:205-209, 1980.
[Context Link]
32. Werner JC, Fripp RR, Whitman V: Evaluation of pediatric surgical patients with congenital heart disease. Surg Clin North Am 63: 1003-1015, 1983.
33. Wise BL, Gleason CA: CT-directed stereotactic surgery in the management of brain abscess. Ann Neurol 6:457, 1979. Bibliographic Links [Context Link]
34. Wispelwey B, Dacey RG Jr, Scheld WM: Brain abscess, in Scheld WM, Whitley RJ, Durack DT (eds): Infections of the Central Nervous System. New York,
Raven Press, 1991, pp 457-486. [Context Link]
35. Yang SY: Brain abscess: A review of 400 cases. J Neurosurg 55:794-799, 1981. Bibliographic Links [Context Link]
36. Yang SY: Brain abscess associated with congenital heart disease. Surg Neurol 31:129-132, 1989. Bibliographic Links [Context Link]
37. Yang SY, Zhao CS: Review of 140 patients with brain abscess. Surg Neurol 39:290-296, 1993. Bibliographic Links [Context Link]
38. Zeidman SM, Geisler FH, Olivi A: Intraventricular rupture of a purulent brain abscess: Case report. Neurosurgery 36:189-193, 1995. Ovid Full Text
COMMENTS
The authors evaluated a large series of patients with brain abscesses related to congenital cyanotic heart disease. Their finding of an increased rate of poor
outcome associated with intraventricular abscess rupture is important and should emphasize the necessity for early diagnosis and treatment of brain
abscesses. Interestingly, the authors recommend only general antibiotic care for patients with abscesses smaller than 2 cm in diameter as revealed by imaging.
I think that a specific bacteriological diagnosis would be facilitated by specific antibiotic therapy and that this would translate into a better results. Thus, we
recommend a stereotactic biopsy or aspiration of all patients with suspected brain abscesses, regardless of size (1). Although full spectrum coverage could be
obtained using multiple antibiotic agents, some drugs are associated with specific morbidities that might be avoided in this fragile patient group. Additionally,
because intraventricular rupture was associated with a poor result, our goal would be to identify the abscess early, diagnose and treat it specifically, and do our
best to prevent abscess enlargement and ventricular rupture.
This is clearly a heterogeneous population of patients with dierent congenital disorders, dierent types of brain abscesses, dierent treatment regimens, and
distinct outcomes. It is dicult to synthesize a management formula because of this, but the authors have performed a service by evaluating this large series.
Douglas Kondziolka
Pittsburgh, Pennsylvania
1. Kondziolka D, Duma C, Lunsford LD: Factors that enhance the likelihood of successful stereotactic treatment of brain abscesses.Acta Neurochir (Wien)
127:85-90, 1994. Bibliographic Links [Context Link]
Takeshita et al. provide a comprehensive review of their large series of 62 patients with brain abscesses associated with cyanotic heart disease managed in the
computed tomography (CT) era. Their treatment philosophy consists of immediate parenteral antibiotic administration with broad spectrum coverage and
surgical aspiration of abscesses larger than 2 cm in diameter causing significant mass eect. For abscesses with intraventricular rupture, parenteral and
intrathecal antibiotic administration was undertaken along with ventricular drainage and aspiration of abscesses larger than 2 cm in diameter, whereas
abscesses smaller than 2 cm in diameter were treated using cerebrospinal fluid drainage and the administration of intrathecal and intravenous antibiotics.
10 of 12
1/24/16, 10:44 AM
Of the 62 patients, 22 patients (35.5%) had multiple abscesses and 21 patients (33.9%) had intraventricular rupture of the abscesses. Poor outcome(totally
disabled or death) occurred in nine patients (14.5%). Multivariate analysis determined that a poor neurologic status, older age, and presence of intraventricular
rupture were the only factors significantly associated with poor outcome. No significant association was determined between abscess size, multiplicity,
location, abscess stage, type of organism, or treatment(excision versus aspiration versus conservative) and outcome. The patients treated conservatively had
small abscesses, usually without significant mass eect, thereby making any outcome comparison with the surgically treated patients invalid. Patients with
multiple abscesses were more frequently immunocompromised with a higher risk of intraventricular rupture when compared with patients with solitary
abscesses. The risk of intraventricular rupture was also higher in abscesses in deeper locations compared to those in more superficial locations, and,
interestingly, of the abscesses located in the periventricular region, those located near the occipital horn were more prone to intraventricular rupture compared
to the abscesses located near the frontal horn. The authors appropriately emphasize the importance of aggressive surgical treatment of abscesses more prone
to intraventricular rupture, namely those larger than 2 cm in diameter with deep locations, especially near the occipital horns, to prevent this morbid
complication.
Interestingly, despite the authors' aggressive surgical approach, a remarkably high number of the patients had intraventricular rupture, which is contradictory to
previous reports, even of patients with cyanotic heart disease. Nevertheless, the surgical morbidity was low (6.5%) and the outcome remarkably good. This
essentially reflects the eectiveness of the authors' treatment paradigm, especially in the setting of complex patients with associated cyanotic heart disease
with an increased risk of cardiopulmonary compromise, coagulation defects, and an immunocompromised state.
Our treatment philosophy also favors an aggressive surgical stance with stereotactic aspiration of all brain abscesses larger than 2 to 2.5 cm in diameter and
repeated aspirations if the size of a smaller abscess increases to greater than this size limit despite antibiotic treatment and/or previous aspiration. Patients
with associated intraventricular rupture of the abscesses also undergo ventricular drainage with intraventricular and parenteral antibiotic administration.
Cerebrospinal fluid drainage is not performed, even for smaller abscesses, mainly because of the risk of herniation. Although broad spectrum antibiotic
coverage is usually indicated in these patients, especially with a poor rate of organism isolation, we restrict the urge to initiate empiric antibiotic treatment
especially in surgical candidates until immediately after the aspiration to minimize the chance of a "sterile" abscess.
The authors present a good analytical report. I hope that with the insights provided herein, low morbidity will become a universal standard in the management
of this complex disease.
Rohit K. Khanna
Mark L. Rosenblum
Detroit, Michigan
The authors provide us with a follow-up report of 62 patients with cyanotic heart disease and brain abscesses. Their original report was published in 1983 and
addressed this situation essentially in the pre-CT era. This is certainly an extraordinarily large experience with this problem and serves as an excellent point of
reference for anyone dealing with patients presenting with brain abscesses and a history of cyanotic heart disease.
It is of interest that one-third of these patients had intraventricular rupture of their brain abscesses, and this was, as expected, associated with worse outcome.
I agree with the author's recommendation of aggressive drainage of an abscess that is in a location threatening to rupture into the ventricle.
William F. Chandler
Ann Arbor, Michigan
Takeshita et al. describe their continuing experience with the management of brain abscesses in patients with congenital cyanotic heart disease. They
retrospectively studied 103 patients treated at their facility from 1968 through 1995. The authors conclude that the mortality rate declined significantly after the
advent of CT, although intraventricular rupture of the brain abscess remained a catastrophic event. The authors clearly have a large experience with this clinical
entity. Because this series used a variety of individualized treatment approaches, the authors are precluded from making statistically definitive statements
regarding optimal surgical therapy in this setting. Nevertheless, their perspectives are of value to the generally neurosurgical readership.
It may be argued that the improvement in survival observed in the CT era may be a result of improvements in broad spectrum antibiotics, rather than
attributable to advances in imaging or in neurosurgical technique. With respect to technique, the authors' use of lumbar drainage for treatment of
intraventricular rupture of abscesses is well described, although it remains controversial in the literature. It is a pleasure to read a broad clinical experience,
carefully considered, that endeavors to discern how best to practice the art of neurosurgery in this challenging clinical setting.
Robert J. Maciunas
Nashville, Tennessee
Key words: Brain abscess; Cyanotic heart disease; Infection; Mortality
IMAGE GALLERY
Select All
11 of 12
Export Selected to PowerPoint
1/24/16, 10:44 AM
Figure 1
Table 1
Table 2
Figure 2
Table 4
Table 3
Table 5
Back to Top
2015 Ovid Technologies, Inc. All rights reserved. OvidSP_UI03.18.00.124, SourceID 69505
12 of 12
About Us
Contact Us
Terms of Use
1/24/16, 10:44 AM

Ovid - Current Treatment of Brain Abscess in Patients With Congenital Cyanotic Heart Disease.

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Ovid - Current Treatment of Brain Abscess in Patients With Congenital Cyanotic Heart Disease.

Încărcat de

Drepturi de autor:

Formate disponibile

Ovid: Current Treatment of Brain Abscess in Patients with C...