J Neurosurg (2 Suppl Pediatrics) 107:119–125, 2007

Benign extracerebral fluid collections in infancy: clinical
presentation and long-term follow-up
LESLIE C. HELLBUSCH, M.D.
Department of Surgery (Neurosurgery), Nebraska Medical Center, Omaha, Nebraska
Object. Benign extracerebral fluid collections are common in infants, but there are unanswered questions regarding
presentation and long-term outcome. This study was undertaken to establish head size at birth, head size at presentation, head growth over time, history of presentation, indications for surgery, and long-term results.
Methods. Cases of benign extracerebral fluid collections in infancy were retrospectively reviewed. Data pertaining
to clinical presentation, progress, and long-term follow-up were evaluated. Patients were divided into two groups on
the basis of the presence or absence of increased signal intensity of fluid in the extracerebral space on T1-weighted magnetic resonance images. Group 1 consisted of 39 patients who had benign extracerebral fluid collections without any
subdural collections. Group 2 consisted of nine patients who had a chronic subdural hematoma or hygroma without a
history of trauma and had generous subarachnoid spaces and/or a history of premature birth. Patients were observed for
an average of 49 months.
Results. Group 1 included 39 patients (ages at presentation 3–12 months); the most common presentation in this
group was macrocephaly (28 cases). Of those Group 1 cases in which data regarding gestational age at birth were available, 15 births were either severely or mildly premature and 14 were full term. Of those Group 1 patients with adequate
data, 14 had an occipitofrontal circumference (OFC) in the 50th to 98th percentile at birth and two had OFCs greater
than the 98th percentile. Measurements obtained at 24 months of age were available for 18 Group 1 patients: OFC was
at approximately the 98th percentile in eight (all eight born at full term) and more than 1 cm larger than the 98th percentile in 10. Head growth in Group 1 patients continued to decrease, approaching the 98th percentile after 24 months
of age. Only three of the 39 patients in Group 1 required shunt placement. Of the 33 Group 1 patients for whom longterm follow-up was available, 30 were developing normally, and three had mild developmental delays.
Group 2 included nine patients (ages at presentation 3–7 months); the most common presentations were large head
(three patients) and seizures (three patients). Data concerning gestational age at birth were available in eight Group 2
cases: birth was severely premature in four and gestation was full term in four. Five Group 2 patients were treated surgically—two with subduroperitoneal (SDP) shunt placement, two with subdural punctures, and one with subdural punctures and SDP shunt placement. Adequate follow-up information was available in eight of nine Group 2 cases; all eight
patients were found to be developing normally.
Conclusions. Group 1 patients usually had a normal head size at birth, they rarely required shunt placement, their head
growth eventually leveled off toward the 98th percentile line, and their outcomes were generally good. Although five of
nine Group 2 patients required subdural punctures and/or SDP shunt placement, that group of patients also developed
well. Infants with nontraumatic subdural hematomas or hygromas, presumably associated with benign extracerebral fluid
collections, can often experience significant resolutions of the hematoma or hygroma within several months without surgical treatment. (DOI: 10.3171/PED-07/08/119)

KEY WORDS • external hydrocephalus • extracerebral fluid collection •
pediatric neurosurgery • subarachnoid space

B

ENIGN extracerebral fluid collection in infancy is also

known as external hydrocephalus. Other names for
this condition include benign extraaxial collection of
infancy,7 benign subdural effusions in infants,16 subdural hygroma,3 pseudohydrocephalus,30 benign communicating hyAbbreviations used in this paper: CSF = cerebrospinal fluid; CT =
computed tomography; OFC = occipitofrontal circumference; MR =
magnetic resonance; SDP = subduroperitoneal; VP = ventriculoperitoneal.

J. Neurosurg: Pediatrics / Volume 107 / August, 2007

drocephalus,12 and extraventricular obstructive hydrocephalus.23 The disorder is characterized by excess fluid, usually
CSF, in the subarachnoid spaces. Unanswered questions related to this condition are investigated in this study. It is
known that a large head circumference (. 98th percentile)
is observed in most cases, but is it observed in all cases? Was
the infant’s head abnormally large at birth, or did it become
large later and at what age? Some patients have been treated with subdural shunts; what were the reasons for shunt
placement? In cases of subdural hematomas involving pa119

L. C. Hellbusch
tients with no known history or examination findings of
trauma, has there been evidence of benign extracerebral
fluid collections on neuroimages? What are the long-term
results of cases involving infants who present with benign
extracerebral fluid collections and how do they compare
with those of cases involving infants who have hematomas
or hygromas and no evidence of trauma?

cessive images (but not until the fluid collection had resolved). Serial images were not obtained unless the patient’s
OFC increased further above the growth curve or developmental milestones were not met. Group 2 patients were followed up until MR imaging or CT scanning demonstrated
resolution of the hematoma/hygroma. Long-term follow-up
of all patients was attempted by telephone.

Clinical Material and Methods
All cases in my clinical records involving patients who
presented between June 1983 and October 2003 were evaluated for inclusion in this study. The records were reviewed
for the following terms: benign extraaxial collection of infancy, benign subdural effusion (in infants), subdural hygroma, pseudohydrocephalus, benign communication hydrocephalus, extraventricular obstructive hydrocephalus,
and external hydrocephalus. Cases involving patients described as having had an enlarged head without hydrocephalus were also included. Patients who presented before August 2002 were selected based on a review of my clinical
records. After August 2002, cases were identified for possible inclusion in the study as the patients presented.
Patients were excluded from the study if they had a
known history of physical trauma, if accidental trauma was
suspected, or if they had other significant disease. Fiftyeight patients were excluded. Sixteen of the 58 excluded patients had a history of accidental trauma, and another seven
had suffered physical trauma through abuse. No patient was
excluded solely due to retinal hemorrhages. Hydrocephalus
was present in 14 cases, and there was a history of meningitis in four. Three cases were excluded because they involved
infants who had been born prematurely and had germinal
matrix hemorrhage, and two cases were excluded because
of craniosynostosis. Another 12 cases were excluded for the
following reasons: inadequate information (two cases); major developmental delays, which were thought to be due
to preexisting cerebral atrophy (two cases); and immunodeficiency, previous cranial surgery (two cases), a major
metabolic problem, osteogenesis imperfecta, macrocephaly
without extracerebral fluid accumulation, DiGeorge syndrome/congenital heart disease, autoimmune enteropathy
(one case each).
The remaining cases were divided into two groups. Group
1 included cases involving patients who presented in infancy
with an OFC greater than the 98th percentile or in whom a
benign extracerebral fluid collection had been diagnosed
without any increased signal intensity of fluid on T1-weighted MR images. Imaging criteria for the diagnosis included
excess fluid in the bifrontal subarachnoid space as evidenced
by a mean distance of more than 4 mm between various frontal gyri and the skull. Group 2 included patients who had presented as infants with a subdural hematoma or hygroma,
generous subarachnoid spaces in addition to the subdural hematoma or hygroma or severe prematurity as in one patient
(Case 4), and no history of trauma.
Follow-up of Group 1 patients was defined as adequate if
1) the patient was followed up at least once 3 months after
the initial diagnosis and was found to be meeting developmental milestones and his or her OFC was not increasing
above growth lines or 2) serial imaging studies were performed (usually at 3-month intervals) until the level of extracerebral fluid was found to have stabilized between suc-

Results
Thirty-nine patients met the criteria for Group 1 (27 boys
and 12 girls). Their age at presentation ranged from 3 to 12
months (median 6.5, mean 7 months). Nine patients (five
boys and four girls) met the criteria for Group 2. Their age
at presentation ranged from 3 to 7 months (median 5
months, mean 5 months).
The most common sign or symptom in Group 1 was macrocephaly, which was present in 28 cases. Vomiting and
fever were each present in four cases. The following were
noted as the chief complaints in one case each: seizure,
“sick” (mastoiditis), “meningitis or hydrocephalus,” twitching, misshaped head, “needs shunt.” Signs and symptoms in
the Group 2 cases included large head circumference (three
cases) and seizures (three cases). Large pupil, vomiting, fussiness, crying, and clogged tear ducts were each present in
different individual cases.
Birth had been severely premature (, 35 weeks gestation) in six cases and mildly premature (35–38 weeks gestation) in nine cases in Group 1. Fourteen Group 1 patients
were born at full term. Birth was severely premature in four
Group 2 cases and full term in four. Information regarding
gestational age at birth was not available in 10 Group 1
cases and one Group 2 case. One Group 2 patient had only
borderline enlargement of the subarachnoid space (3–4 mm)
but was included in the study because of premature birth.
At birth, most Group 1 patients with benign extracerebral fluid collections had an OFC in the 50th to 98th percentile (Figs. 1 and 2). Fourteen infants who were not born
prematurely for whom information was available had
OFCs at birth in the 50th to 98th percentile, two had OFCs
at birth less than the 50th percentile, and one had an OFC
greater than the 98th percentile. When patients were evaluated at 24 months of age, approximately half of all those
who had not been born prematurely and for whom information was available (eight) had an OFC around the 98th
percentile, and the remaining patients in that subgroup (10)
had an OFC more than 1 cm larger than the 98th percentile.
In this study, only three of the 39 patients in Group 1 required shunt placement (Table 1). Two patients received
SDP shunts—one because of a benign extracerebral fluid
collection, a markedly enlarged head, and the development
of subdural hematoma/hygroma (Case 1); the other because
of an OFC much greater than 98th percentile and some
vomiting. One patient received a VP shunt. The patient who
received a VP shunt (Case 2) had a markedly increased
OFC, a significant benign extracerebral fluid collection, and
enlarged ventricles. Five of the nine Group 2 patients were
treated surgically: two with SDP shunt placement, two with
subdural taps only, and one with SDP shunt placement and
subdural taps (Table 1).
A follow-up CT or MR imaging study was obtained in
17 of 39 Group 1 patients, and in eight of nine Group 2.
The most recent information available for most patients

120

J. Neurosurg: Pediatrics / Volume 107 / August, 2007

Benign extracerebral fluid collections in infancy
TABLE 1
Surgical procedures performed

FIG. 1. Head growth chart from a Group 1 case illustrating an
OFC between the 50th and 98th percentiles at birth, a finding that
was characteristic of 14 infants who were not born prematurely, and
an OFC near the 98th percentile at 24 months, characteristic of eight
Group 1 children who were not born prematurely.

was obtained by telephone interview with a parent. The
long-term outcomes for those patients were very good. Of
the Group 1 patients, 33 had adequate follow-up (range
7–129 months after initial diagnosis, mean 49 months, median 36 months). Thirty Group 1 patients displayed development within normal limits. Two were found to have mild
delay at 26 and 78 months after diagnosis, and one had
speech delay at 7 years of age (Case 1). Eight of the nine
Group 2 patients, were found to be developing within normal limits. Follow-up was inadequate in the remaining case
(Table 2).

Group No. & Type of Procedure

No. of Patients

Group 1
SDP shunt placement
VP shunt placement
Group 2
SDP shunt placement
subdural taps
subdural taps & SDP shunt placement

39
2
1
9
2
2
1

Initial images obtained in eight Group 2 patients (three of
whom were born $ 6 weeks prematurely) showed both subdural hematoma/hygroma and benign extracerebral fluid
collections. Also, the initial image obtained in one patient
(born at 33 weeks’ gestation) showed only subdural hematoma/hygroma, with no extracerebral fluid collection. In one
Group 1 patient (Case 1), the initial image showed a benign
extracerebral fluid collection, and a later image showed subdural hematoma/hygroma. In three of the eight Group 2 cases in which the initial images showed both benign extracerebral fluid collections and subdural hematoma/hygroma,
follow-up studies showed an increase in the subdural hematoma/hygroma before successful treatment.
Illustrative Cases
Case 1. This 4.5-month-old Group 1 male infant presented with an enlarging head circumference (OFC 55.5 cm, .
98th percentile) and no history of trauma. The initial MR
images showed benign extracerebral fluid collections (Fig.
3); follow-up images obtained when the child was 12
months old showed significant subdural hematoma/hygroma (Fig. 3C). The hematoma/hygroma was treated with a
right SDP shunt. At 40 months of age, the patient’s OFC
was still above the 98th percentile, but follow-up MR
images showed less extracerebral fluid, and the child’s OFC
had increased only 1 cm in the preceding 2 years. At the
most recent follow-up, the child was 7 years of age, developing normally, and attending regular school. He was receiving speech therapy for articulation at school because of
what appeared to be a hereditary problem. (His mother and
other relatives received speech therapy for the same reason.)
Case 2. This 10-month-old Group 1 male infant had an increased OFC and a benign extracerebral fluid collection was
diagnosed. He was born prematurely at 27 weeks of gestational age. A CT scan showed a moderate amount of extracerebral fluid and mild to moderate ventriculomegaly (Fig.
4 left). Over the course of three cranial ultrasound examTABLE 2
Long-term results in cases in which follow-up was adequate*

FIG. 2. Head growth chart from a Group 1 case illustrating an
OFC at 24 months that was more than 1 cm greater than the 98th
percentile, a finding that was characteristic of 10 patients who were
not born prematurely.

J. Neurosurg: Pediatrics / Volume 107 / August, 2007

Group & Outcome

No. of Patients

Group 1
within normal developmental limits
speech delay at age 84 mos
mild developmental delay at 26 & 78 mos
Group 2
within normal developmental limits

33
30
1
2
8
8

* Follow-up was inadequate in six Group 1 cases and one Group 2 case.

121

L. C. Hellbusch

FIG. 4. Case 2. Computed tomography scans. Left: Scan obtained at 10 months of age showing extracerebral fluid and mild to
moderate ventriculomegaly. Right: Scan obtained at 15 months of
age after placement of VP shunt showing no initial improvement in
the amount of bifrontal extracerebral fluid.

FIG. 3. Case 1. Axial T1-weighted MR images. A: Image obtained at 6 months of age showing wide bifrontal extracerebral
spaces. B: Image obtained at 6 months of age; the extracerebral
fluid extends down to the anterior temporal areas. C: Image obtained at 1 year of age illustrating significant subdural hematoma/
hygroma. D: Image obtained at 40 months of age showing marked
resolution of benign extracerebral fluid collection and subdural hematoma/hygroma following SDP shunt placement.

inations, the ventricular measurements were found to have
increased slightly. The patient’s initial OFC was 53 cm (.
98th percentile), and when he was 14 months old it had increased to 55 cm. At 15 months of age, the patient underwent VP shunt placement because his OFC was continuing
to increase. At a follow-up visit when he was 22 months
old, his OFC was found to have remained stable since the
shunt placement (Fig. 4 right). At that time he was developing well, but he was subsequently lost to follow-up because
the family moved across the country.
Case 3. This 6-month-old Group 2 male infant presented
with one pupil larger than the other and no other symptoms.
The child was born prematurely after 27 weeks of gestation.
Ultrasound and MR imaging examinations demonstrated a
benign extracerebral fluid collection and a left subdural hematoma/hygroma (Fig. 5A). The MR images showed mild
ventriculomegaly and revealed that the ventricle size had
increased since the ultrasound examination. The left subdural hematoma was 8 mm in width and covered the middle
and superior aspects of the left cerebral hemisphere. The
OFC was 40.5 cm when the patient was 6 months of age
and 42.5 cm at 7 months of age. The MR images obtained
at 7 months of age showed resolution of the subdural hematoma, and the infant did not appear to be experiencing any
symptoms at that time (Fig. 5C). The patient was doing
well with no intervening therapy at a follow-up visit at 8
months of age and was still developing well at the age of 48
months.
Case 4. This 5-month-old Group 2 female infant, a twin
122

who was born after 33 weeks of gestation, had a history of
vomiting, fussiness, and crying. The initial MR imaging
study showed a bifrontal increased signal, which represented excessive fluid, and the anterior fontanelle was moderately tense (Fig. 6 left). The OFC at 5 months was 42.5 cm.
The infant underwent insertion of a right SDP shunt because
of persistent vomiting. Microscopically, some red blood
cells were noted in the xanthochromic CSF. At 8 months of
age, the patient was doing well, and MR images obtained at
this time showed a decrease in the fluid volume (Fig. 6
right). When the patient was 16 months old, the OFC was
48 cm, and the anterior fontanelle had decreased to 1 cm in
diameter and was flat. There was no evidence of any collection of extracerebral fluid, and the patient was still doing
well. At 4 years of age, the patient had completed her 1st
year of preschool and was developing normally socially and
intellectually. She participated in a study performed by students at a local university in which motor skills were evaluated using a battery of tests (such as stacking blocks), and
her performance was above average for her age group.
Discussion
Fukuyama and colleagues10 reported on the normal CT
appearance of the subarachnoid space. It was found that the
growth of the brain and skull are virtually parallel until the
age of 3 months and that there is barely any subarachnoid
space. Throughout the 1st year of life, imbalance was found
between the growth of the skull and the brain, but this eventually reached equilibrium after 1 year. The subarachnoid
space was considered abnormal if it exceeded 8 mm in
width in the 1st year and abnormal in the 2nd year if it exceeded 4 mm. Prassopoulos and coauthors26 reported the
maximum frontal CSF compartment width to be 4 mm, with
the abnormal range being 6 to 23 mm. Thus opinion differs
regarding the criteria for an abnormal accumulation of fluid.
Also, there are several ways the measurement can be taken
with similar, but not necessarily identical, results. Should the
measurement be taken from the frontal gyrus farthest from
the skull or should it represent an average distance from various gyri to the skull? In the cases reported in this paper, the
latter technique was used.
J. Neurosurg: Pediatrics / Volume 107 / August, 2007

Benign extracerebral fluid collections in infancy

FIG. 5. Case 3. Axial T1-weighted MR images. A: Image obtained at 5 months of age showing extracerebral fluid
and left subdural hematoma/hygroma approximately 8 mm in width. B: Image obtained at 5 months of age showing the
extracerebral fluid extending down to the anterior temporal region bilaterally. C: Follow-up image obtained at 7 months
of age demonstrating spontaneous resolution of the hematoma/hygroma.

The most common physical characteristic of external
hydrocephalus is a large head. In the present study, 28 of 39
patients in Group 1 presented with macrocephaly. In Group
2, three of nine patients had enlarged heads. Alvarez et al.2
reported a family history of macrocephaly in 88% of idiopathic external hydrocephalus cases. Castro-Gago et al.8
identified a history of macrocephaly in one of the progenitors in 38.46% of cases. Nevertheless, Akaboshi et al.1 reported a case of benign extracerebral fluid collection in a
premature infant with autosomal dominant microcephaly.
Other characteristics of benign extracerebral fluid collection
in infancy include full fontanelles and enlarged ventricles.16
A wide frontal subarachnoid space is another obvious characteristic of this condition, in contrast to cerebral atrophy, in
which the widened subarachnoid space is more generalized.
Maytal et al.14 differentiated external hydrocephalus from
cerebral atrophy because cerebral atrophy showed prominent sulci throughout without a disproportionate widening
of the bifrontal subarachnoid space. In all the cases reported in this study, the patients had bifrontal and not diffusely
prominent sulci.
Andersson et al.3 reported on nine patients with widened
subarachnoid spaces. Of these nine, seven underwent craniotomies and two underwent shunt placement. In each of the

FIG. 6. Case 4. Magnetic resonance images. Left: Axial T1weighted image obtained at 5 months of age illustrating increased
bifrontal extracerebral spaces with slightly increased signal intensity indicating a probable hematoma or hygroma. Right: Axial T2weighted image obtained at 8 months of age illustrating a decrease
in extracerebral fluid following SDP shunt placement.

J. Neurosurg: Pediatrics / Volume 107 / August, 2007

patients who underwent craniotomies, a “deep” arachnoid
space was found. Three of those seven required a VP shunt
to control CSF leakage following their operations. Kapila et
al.11 looked at CT scans obtained in infants who had enlarged subarachnoid spaces. They found 16 patients to have
associated subdural hematomas, which were documented
by CT (seven patients), subdural tap (one patient), or both
CT and subdural tap (eight patients). The authors deduced
that the enlarged subarachnoid spaces predisposed those
infants to subdural hematomas because clinical follow-up
showed no evidence of cerebral atrophy. Orrison et al.21 described CT findings in six cases of chronic subdural hematoma (effusion) in infancy, but they did not differentiate well
between chronic subdural hematoma, chronic subdural effusion, or simple CSF accumulation (as in external hydrocephalus), despite subdural punctures, which should have
allowed better definitions of these entities. Current MR imaging techniques permitted differentiation between hematoma/hygroma and CSF in all of the cases reported in the present study.
Infants with benign extracerebral fluid collections can
have subdural hematoma/hygroma complications with slight
trauma or none. Ravid and Maytal28 suggested that in some
infants with benign extracerebral fluid collections, a subdural hematoma may develop with minimal or no trauma due to
“stretching of the bridging veins in the unusually widened
subarachnoid spaces.” This stretching of the bridging veins
leaves the veins without a support system, which predisposes them to bleeding with little or no trauma. Papasian et al.22
developed a model of the intracranial space that can predict
situations in which infants with benign extracerebral fluid
collections will develop extraaxial hemorrhage when healthy infants will not. Their equations predict an increased incidence of stretch injuries to veins in the extraaxial space in
patients with benign extracerebral fluid collections, findings
that correlate with those of Ravid and Maytal.28 Laubscher
and colleagues13 studied the cases of two infants in whom
chronic subdural hematomas developed, one without trauma
and the other after a 40-cm fall. These infants had large
heads (one in the 90th–97th percentile and the other above
the 97th percentile). He also described a third infant in
whom a subdural hygroma developed spontaneously at the
age of 9 months. A CT scan obtained in this patient showed
enlarged subarachnoid spaces. The authors postulated that
benign extracerebral fluid collections could predispose in123

L. C. Hellbusch
fants to spontaneous subdural hematoma. They also suggested that benign extracerebral fluid collections in infancy
represent a variant of “primitive megalencephaly.” Our
Group 2 patients were thought to have extracerebral hematoma/hygroma despite a history of slight trauma or none.
Imaging studies showed some evidence of benign extracerebral fluid collections along with the hematoma/hygroma in
eight of these nine patients. The excellent outcome of eight
of the nine cases (one patient was lost to follow-up) is further evidence of the absence of serious trauma in this group.
This study does not answer the question regarding whether
infants with benign extracerebral fluid collections can have
retinal hemorrhages without apparent trauma.
Mori and associates17 described three infants with extracerebral fluid collections who developed subdural hemorrhages. One of the infants had a fall of 50 cm, and subdural
punctures caused another infant’s subdural hemorrhage.
The third infant did not have a history of trauma. A unilateral subdural hematoma developed in this infant and was
treated with bur-hole drainage. One year later, the patient
developed a subdural hematoma on the contralateral side.
The authors quoted Dr. Rekate as saying that a subdural hematoma can easily develop as a consequence of head shaking in infants who have a subarachnoid fluid collection.
Azais and Echenne5 found that five of 41 patients in whom
“benign enlargement of the subarachnoid spaces” was demonstrated on ultrasonographic examinations and CT studies
had “spontaneous” subdural hemorrhage. Pittman25 regarded the presence of a subdural hematoma by itself, in the absence of other evidence of inflicted injury, in a child with
external hydrocephalus insufficient to prove abuse.
Premature birth might predispose infants to the development of subarachnoid fluid collections.17 Fifteen of the
Group 1 patients and four of the Group 2 patients were
known to have been born prematurely.
With MR imaging technology, it has become easier to diagnose benign extracerebral fluid collections in infancy accurately and to differentiate this condition from subdural
hematoma or hygroma than it was previously. DeVries et
al.9 reported on the differentiation between subdural effusions and external hydrocephalus using ultrasonography
and MR imaging. Aoki4 reported on the role of MR imaging in distinguishing between subdural effusion and subarachnoid space enlargement and stated, “all patients with
subdural effusion had fluid that was greater in intensity on
at least one sequence using T1-weighted, proton-density,
and T2-weighted MRI images. The flow void sign, indicating vessels in the fluid spaces, was not seen in any of these
eight patients.”
Infantile subdural hematoma/hygroma can decrease without surgical treatment. Hematoma/hygroma can decrease
fairly rapidly, within 1 or 2 months. Ravid and Maytal28 stated that the prognosis of patients with external hydrocephalus and subdural hematoma in infancy is usually better than
that of other patients with subdural hemorrhage and also
stated that in most infants the hematoma will resolve spontaneously. They added three cases of their own, in two of
which the hematomas resolved spontaneously and in one of
which surgical drainage was performed despite the patient’s
being asymptomatic. In four of nine of the Group 2 cases in
the present study the hematoma/hygroma resolved spontaneously.
Mori and coauthors16 reported on 20 cases of benign sub124

dural effusion in infants; 20 of the infants developed essentially normally without surgery. Nickel and Gallenstein18
followed the development of nine infants who had enlarged
subarachnoid spaces. None of these infants developed hydrocephalus, and the anomaly had a normal prognosis. Robertson et al.29 observed six infants who had subdural collections of fluid, and in five of the six there were normal
long-term prognoses. Ment et al.15 reported that 11 of 12 infants with enlarged subarachnoid spaces developed normally, and these authors proposed on the basis of the findings
of short-term follow-up examinations that benign enlargement of the subarachnoid spaces is rarely associated with
neurological dysfunction.
Castro-Gago et al.8 conducted a retrospective study involving 39 patients with benign idiopathic external hydrocephalus. Age at diagnosis ranged from 1.33 to 25 months
(mean 8.4 months), the mean age at which clinical control
was obtained was 3.36 years, and the process was observed
to resolve in 14 cases. In five cases, there was motor retardation, and one patient displayed psychomotor retardation.
In the present study, there were no cases of moderate or severe developmental delay in either Group 1 or Group 2,
perhaps because patients with coexisting disease processes
were excluded from the study.
The precise cause of benign extracerebral fluid collections in infancy has not yet been determined although several theories have been proposed. Barlow6 stated that the
condition is caused by anatomical or functional obstruction
of various causes at the level of the arachnoid villi. Mori et
al.16 hypothesized that the dilation of the subarachnoid space
may be caused by communicating hydrocephalus due to a
distal (high convexity or parasagittal) block. In defining external hydrocephalus, Raimondi27 theorized that there is a
free movement of CSF from the subarachnoid space through
the membrane into the subdural space in the presence of a
pathological accumulation of CSF. Raimondi implied that
both compartments contained CSF. Odita20 proposed that the
widening of the subarachnoid space in children might be a
variation of normal brain development, with a transient accumulation of CSF in the frontal region. Therefore, according to Odita, any child with a rapidly increasing head size
may have an “accumulation of this widening.”20 The results
of the present study showed that a fairly high percentage of
patients in Group 1 (15 of 29 for whom information on gestation was available) and in Group 2 (four of eight) were
born prematurely and therefore might experience rapid
catch-up head growth. Nogueira and Zaglul19 proposed that
cephalocranial disproportion is the basic underlying entity.
Piatt24 hypothesized that external hydrocephalus is caused
by a developmental disturbance of skull growth—an excessive response of the calvaria signals that link its growth to
the growth of the brain. A transient disturbance of CSF circulation, possibly due to delayed maturation of the arachnoid granulations, is commonly accepted as a cause of benign extracerebral fluid collections in infancy.
Conclusions
Both groups of cases in this series generally had good
outcomes, and shunt placement was rarely required in
Group 1. It is still difficult to differentiate between cases of
mild benign extracerebral fluid collections and normal infant head size. Nevertheless, extracerebral fluid collections
J. Neurosurg: Pediatrics / Volume 107 / August, 2007

Benign extracerebral fluid collections in infancy
have become easier to diagnose accurately and to differentiate from subdural hematomas and hygromas with advances in MR imaging technology. Patients who have benign extracerebral fluid collections in infancy usually have
an OFC in the 50th to 98th percentile at birth, and at 24
months OFC is in the 98th percentile in about half of the
cases and more than 1 cm greater than the 98th percentile
in the rest. In infants with subdural hematomas or hygromas
with no history of trauma, the lesions can decrease within 1
to 2 months without surgical treatment, and infants with benign extracerebral fluid collections can have subdural hematoma/hygroma complications with little or no trauma.
Acknowledgment
I thank Rachel Meyer for her editorial guidance in preparing this
paper.
References
1. Akaboshi I, Ikeda T, Yoshioka S: Benign external hydrocephalus in a boy with autosomal dominant microcephaly. Clin Genet 49:160–162, 1996
2. Alvarez LA, Maytal J, Shinnar S: Idiopathic external hydrocephalus: natural history and relationship to benign familial macrocephaly. Pediatrics 77:901–907, 1986
3. Andersson H, Elfverson J, Svendsen P: External hydrocephalus
in infants. Childs Brain 11:398–402, 1984
4. Aoki N: Extracerebral fluid collections of infancy: role of magnetic resonance imaging in differentiation between subdural effusion and subarachnoid space enlargement. J Neurosurg 81:
20–23, 1994
5. Azais M, Echenne B: [Idiopathic pericerebral swelling (external
hydrocephalus) of infants.] Ann Pediatr (Paris) 39:550–558,
1992 (Fr)
6. Barlow CF: CSF dynamics in hydrocephalus—with special attention to external hydrocephalus. Brain Dev 6:119–127, 1984
7. Carolan PL, McLaurin RL, Towbin RB, Towbin JA, Egelhoff
JC: Benign extra-axial collections of infancy. Pediatr Neurosci
12:140–144, 1985–1986
8. Castro-Gago M, Perez-Gomez C, Novo-Rodriguez MI, BlancoBarca O, Alonso-Martin A, Eiris-Punal J: [Benign idiopathic
external hydrocephalus (benign subdural collection) in 39 children: its natural history and relation to familial macrocephaly.]
Rev Neurol 40:513–517, 2005 (Sp)
9. De Vries LS, Smet M, Ceulemans B, Marchal G, Wilms G, de
Roo M, et al: The role of high resolution ultrasound and MRI
in the investigation of infants with macrocephaly. Neuropediatrics 21:72–75, 1990
10. Fukuyama Y, Miyao M, Ishizu T, Maruyama H: Developmental
changes in normal cranial measurements by computed tomography. Dev Med Child Neurol 21:425–432, 1979
11. Kapila A, Trice J, Spies WG, Siegel BA, Gado MH: Enlarged
cerebrospinal fluid spaces in infants with subdural hematomas.
Radiology 142:669–672, 1982
12. Kendall B, Holland I: Benign communicating hydrocephalus in
children. Neuroradiology 21:93–96, 1981
13. Laubscher B, Deonna T, Uske A, van Melle G: Primitive mega-

J. Neurosurg: Pediatrics / Volume 107 / August, 2007

lencephaly in children: natural history, medium term prognosis
with special reference to external hydrocephalus. Eur J Pediatr 149:502–507, 1990
14. Maytal J, Alvarez LA, Elkin CM, Shinnar S: External hydrocephalus: radiologic spectrum and differentiation from cerebral atrophy. AJR Am J Roentgenol 148:1223–1230, 1987
15. Ment LR, Duncan CC, Geehr R: Benign enlargement of the subarachnoid spaces in the infant. J Neurosurg 54:504–508, 1981
16. Mori K, Handa H, Itoh M, Okuno T: Benign subdural effusion
in infants. J Comput Assist Tomogr 4:466–471, 1980
17. Mori K, Sakamoto T, Nishimura K, Fujiwara K: Subarachnoid
fluid collection in infants complicated by subdural hematoma.
Childs Nerv Syst 9:282–284, 1993
18. Nickel RE, Gallenstein JS: Developmental prognosis for infants
with benign enlargement of the subarachnoid spaces. Dev Med
Child Neurol 29:181–186, 1987
19. Nogueira GJ, Zaglul HF: Hypodense extracerebral images on
computed tomography in children. “External hydrocephalus”: a
misnomer? Childs Nerv Syst 7:336–341, 1991
20. Odita JC: The widened frontal subarachnoid space. A CT comparative study between macrocephalic, microcephalic, and normalcephalic infants and children. Childs Nerv Syst 8:36–39,
1992
21. Orrison WW, Robertson WC, Sackett JF: Computerized tomography in chronic subdural hematomas (effusions) of infancy.
Neuroradiology 16:79–81, 1978
22. Papasian NC, Frim DM: A theoretical model of benign external
hydrocephalus that predicts a predisposition towards extra-axial
hemorrhage after minor head trauma. Pediatr Neurosurg 33:
188–193, 2000
23. Pettit RE, Kilroy AW, Allen JH: Macrocephaly with head growth
parallel to normal growth pattern: neurological, developmental,
and computerized tomography findings in full-term infants. Arch
Neurol 35:518–521, 1980
24. Piatt JH Jr: Monozygotic twins discordant for external hydrocephalus. Pediatr Neurosurg 35:211–215, 2001
25. Pittman T: Significance of a subdural hematoma in a child with
external hydrocephalus. Pediatr Neurosurg 39:57–59, 2003
26. Prassopoulos P, Cavouras D, Golfinopoulos S, Nezi M: The size
of the intra- and extraventricular cerebrospinal fluid compartments in children with idiopathic benign widening of the frontal
subarachnoid space. Neuroradiology 37:418–421, 1995
27. Raimondi AJ: Pediatric Neurosurgery: Theoretical Principles—
Art of Surgical Techniques. Berlin: Springer, 1987, pp 453–491
28. Ravid S, Maytal J: External hydrocephalus: a probable cause for
subdural hematoma in infancy. Pediatr Neurol 28:139–141,
2003
29. Robertson WC Jr, Chun RW, Orrison WW, Sackett JF: Benign
subdural collections of infancy. J Pediatr 94:382–386, 1979
30. Sahar A: Pseudohydrocephalus-megalocephaly, increased intracranial pressure and widened subarachnoid space. Neuropadiatrie 9:131–139, 1978
Manuscript submitted September 1, 2006.
Accepted March 5, 2007.
Address reprint requests to: Leslie C. Hellbusch, M.D., 8005
Farnam Drive, Suite 305, Omaha, Nebraska 68114. email: lhellbusch
@msn.com.

125