Computed
Tomography
of the Head
in the
Evaluation
of Microcephaly
Margie Jaworski, MD, Joseph H. Hersh, MD, Jane Donat, MD,
Loretta T. Shearer, MD, and Bernard Weisskopf, MD
From the Child Evaluation Center, Department of Pediatrics; and the Departments of Neurology and Radiology, University of Louisville, Louisville
ABSTRACT. Eighty-five infants and children found to
have microcephaly had computed tomographic (CT)
brain scans performed. A greater degree of microcephaly correlated with the finding of atrophy or ventricular
dilation on CT scan. Patients who had known preceding destructive brain insults had the highest incidence of
abnormal findings on scans (20/22). Patients who had
CNS dysfunction of unknown etiology had the lowest
frequency ofabnormal findings (12/33); however, in three
of these patients, a previously unsuspected brain
malfor-mation was found on CT scan. Patients who had other congenital anomalies had an intermediate proportion of abnormal findings on CT scans (20/30), and in 11 of these scans, a previously unsuspected or only partly sus-pected brain malformation was diagnosed. Discovering
previously unsuspected information or finding supportive
data regarding the basis for the underlying disease
proc-ess, being able to provide a more specific developmental prognosis and accurate genetic counseling, justifies the
inclusion of a CT scan of the head in the evaluation of the microcephalic child. Pediatrics 1986;78:1064-1069; computed tomography, microcephaly, head circumference.
With the advent of computed tomography (CT), a noninvasive device is now available to evaluate brain parenchyma and ventricular size.’ Recom-mendations for its use in children have included evidence of increased intracranial pressure,
macro-crania, rapid increase in head size, changing or focal
neurologic signs, coma of unknown etiology, and neurocutaneous syndromes and for follow-up after
operative intervention or radiation therapy. Static
encephalopathies such as cerebral palsy, mental
retardation, or seizure disorders are usually not
considered adequate indications because CT
find-Received for publication Dec 8, 1985; accepted Feb 19, 1986. Reprint requests to (J.H.H.) Child Evaluation Center, 334 Broadway, Louisville, KY 40202.
PEDIATRICS (ISSN 0031 4005). Copyright © 1986 by the American Academy of Pediatrics.
ings are unlikely to change the treatment plan.2 To
date, no studies have been published critically
eval-uating the use of the CT scan in patients who have microcephaly. This study was designed to ascertain the types and frequencies of abnormalities seen on the CT brain scan in patients who have
micro-cephaly and to determine whether the information
obtained would be helpful for diagnosis, prognosis, and genetic counseling.
PATIENTS AND METHODS
The study population comprised 85 infants and children who were seen during a 7-year period by
the staff of the Child Evaluation Center in Louis-ville, a multidisciplinary center where children are
evaluated in a genetic and developmental
disabili-ties clinic. Microcephaly was identified in all of the patients and each underwent a CT scan of the head.
The criteria used for microcephaly was a head
cir-cumference 2 SD below the mean,3’4 based on the standards derived by Nelhaus.5 There were 43 male
and 32 female patients. They ranged in age from
neonates to 16 years and 67 patients were younger than 3 years of age.
The majority of CT scans were performed using
an EXEL Elscint 2002 CT body scanner, although,
in a few of the earlier cases, an EMI 5005 CT body scanner was used. In each case, a slice thickness of 1 cm was taken with an average number of nine slices per scan, and all studies were performed
without contrast. Each scan was reviewed by a
pediatric radiologist (L.S.) and a pediatric neurol-ogist (J.D.), without knowledge of the clinical data,
and they came to a consensus regarding the final
reading.
Patients were divided into four groups based on the CT results. Group 1 had normal findings on CT
scans. Group 2 had abnormalities characterized by
had evidence of moderate to severe atrophy and/or ventricular dilation. In group 4, patients did not have evidence of atrophy or ventricular dilation but
were found to have parenchymal abnormalities.
Atrophy was determined by evaluating cerebral sulci, gyri, and interhemispheric fissures.
Ventric-ular size was evaluated for the lateral, third, and fourth ventricles. Parenchymal lesions occurred in groups 2 through 4 and included calcifications, de-creased density, and disruption ofwhite and/or gray matter.
RESULTS
Of the 85 patients seen, 33 (39%) had normal and 52 (61%) had abnormal findings on CT scans. In group 1 (Table 1), which consisted of33 patients
who had microcephaly and normal scan findings, the average head circumference was 3.1 SD below
the mean with a range of 2.05 to 6.6 SD below the
mean. Diagnoses in this group included
develop-mental delay or CNS dysfunction without a history of previous brain injury in 21 (64%) children, CNS dysfunction secondary to hypoxia-ischemia and/or meningitis in two (6%), and a variety of conditions which included recognizable patterns of malforma-tion, nonrandom associations, and isolated
struc-tural malformations in 20 (30%).
Group 2 (Table 1) comprised 26 patients who had mild atrophy and/or ventricular dilation (Fig 1).
The average head circumference was 3.9 SD below
the mean with a range of 2.0 to 8.4 SD below the mean. Nine (35%) patients in this group were also found to have parenchymal abnormalities (Table
2). There were six (23%) patients who had CNS
dysfunction alone, including one who had a paren-chymal lesion, namely, schizencephaly, not
previ-TABLE 1. Diagnostic Categories of Patients With Microcephaly Who Underwent Computed Tomographic Scans of the Head*
Diagnosis Group
Normal Mild Atrophy and/or Moderate to Severe Parenchymal Lesion Alone (n = 33) Ventricular Dilation Atrophy and/or (n = 2)
(n = 26) Ventricular Dilation
(n = 24)
CNS dysfunction of unknown etiology 21 6 6 0
CNS dysfunction secondary to hypoxia- 2 11 9 0
ischemia/intraventricular hemorrhage and/or meningitis
Other etiologies 10 9 9 2
TABLE 2. Parenchymal Lesions Identified by Computed Tomographic Scan in the Three Groups
Group CNS Dysfunction of Unknown CNS Dysfunction Secondary to Other Etiologies
Etiology Hypoxia-Ischemia/Intraventricular Hemorrhage and/or Meningitis
1 (n
=
33)2 (n = 26) Congenital developmental:
schizencephaly*
Acquired:
hypoxia-ischemia-diffuse low density; decreased
attenuation white matter;
por-encephaly, focal atrophy; in-traventricular
hemorrhage-patchy low density;
poren-cephaly
Congenital developmental: lobar
holoprosencephaly (infant
dia-betic mother)*; lobar
holopro-sencephaly (arrhinencephaly) Acquired:
infection-periven-tricular and intraparenchymal
calcifications (intrauterine
in-fection)* 3 (n = 24) Congenital developmental: lobar
holoprosencephaly,* white
matter low-density gyral ab-normalities, agenesis of
cor-pus callosum, posterior fossa cyst
Acquired: hypoxia-ischemia-de-creased attenuation; decreased
attenuation; porencephaly; bi-lateral cerebral infarctions;
low-density areas; hypoxia-is-
chemia/meningitis-poren-cephaly; postnatal trauma-left parietooccipital infarction
Congenital developmental: schizencephaly (fetal alcohol
syndrome); multiple low-den-sity areas (multiple congenital
anomalies); lobar holoprosen-cephaly (holoprosencephaly
se-quence)
Acquired:
infection-periven-tricular calcifications (intra-uterine infection);* other-white matter calcifications (Cockayne syndrome)*
4 (n = 2) Congenital developmental:
pri-mary brain malformation (en-cephalocele)
Acquired:
infection-periven-tricular calcifications (cyto-megalovirus)
Recognizable disorder not suspected prior to obtaining computed tomographic scan.
ously suspected. There were 11 (42%) patients who had a history of hypoxic-ischemic encephalopathy
or intraventricular hemorrhage, and five of those had parenchymal lesions including left frontal por-encephaly, diffuse or patchy low density, left
occip-ital porencephaly and focal atrophy, and decreased
attenuation of white matter in individual patients. Of the remaining nine (35%) patients with other diagnoses including chromosomal abnormalities, various syndromes, and intrauterine viral
infec-tions, there were three with parenchymal lesions. Periventricular and intraparenchymal calcifica-tions and atrophy of the right parietal region were
present in one patient, and the finding assisted in
establishing a diagnosis of an intrauterine infection
which had not been considered clinically. Lobar holoprosencephaly, not suspected in an infant of a diabetic mother with branchial arch and cardiac and renal lesions, occurred in another patient.
Fi-nally, agenesis of the corpus callosum, abnormal
occipital lobes, and poor visualization of the frontal
horns, all suggestive of holoprosencephaly, were
seen in a patient with a clinical diagnosis of
ar-rhinencephaly.
In group 3 (Table 1), which included those with moderate to severe atrophy and/or ventricular
di-lation (Figs 2 and 3), there were 24 patients who
had average head circumferences of 3.8 SD below
the mean, ranging from 2.1 to 8.5 SD below the mean. Of six (25%) patients with CNS dysfunction alone, there were three with parenchymal lesions
including lobar holoprosencephaly not suspected on clinical examination, white matter low-density
gyral abnormalities, and agenesis of the corpus
callosum with a posterior fossa cyst. Of nine
(37.5%) patients having CNS dysfunction after hy-poxic-ischemic encephalopathy, including one in whom neonatal meningitis developed, seven had parenchymal lesions, including low-density areas of white and gray matter in three, cerebral infarctions in two, and porencephaly in two. Of the nine
(37.5%) patients with other diagnoses, which
in-cluded three with neural tube defects, four with a nongenetic basis for the CT findings, and one each
with Cockayne syndrome, holoprosencephaly
se-quence, and multiple congenital anomalies, there were five parenchymal abnormalities (Table 2)
in-cluding calcifications in two patients, one
periven-tricular, and another intraparenchymal, lobar
holoprosencephaly in one, schizencephaly in one,
and diffuse low-density in one. Thus, of 24 patients in group 3, 15 (62%) had defects in the parenchyma
(Table 2). In both patients with intracranial calci-fications, a specific diagnosis was not entertained
prior to obtaining a CT scan. Periventricular
;scan of]
ing moderately dilated lateral and third ventricles. Sulci
over convexities of brain are prominent.
consistent with moderate brain atrophy.
3. scan
ing marked dilation of lateral and third ventricles, in- brain atrophy. creased subarachnoid spaces, and prominent sulci over
and no other clinical abnormalities led to suspicion
of an intrauterine infection as the likely diagnosis for the neurologic impairments and microcephaly.
CT findings characterized by basal ganglia
calcifi-cations in another patient who presented with post-natal growth retardation, hearing loss, and deeply
set eyes, assisted in establishing a diagnosis of
Cockayne syndrome.
Group 4 (Table 1) consisted of two patients who did not have evidence of ventricular dilation or atrophy but were found to have parenchymal changes (Table 2). One of these patients had an
encephalocele and a severe primary brain
malfor-mation. The other patient had periventricular
cal-cifications that supported a clinical diagnosis of an intrauterine viral infection based on intrauterine
growth retardation, purpura, and chorioretinitis.
The overall analysis of variance was not
signifi-cant (P = .12) when the head circumferences were
compared in the three groups. However, a planned
contrast revealed that the mean head circumference of the patients with moderate to severe atrophy, ie group 3, was significantly smaller than that of group 1 which had normal CT scan findings (P = .049);
whereas there was no significant difference between
the mean head circumferences of groups 1 and 2
(normal v mild) (P = .146).
scan of the head is generally based on physical and neurologic findings. Therefore, patients in our
study were also categorized into three groups to determine the significance of a CT scan given a specific clinical diagnosis: patients who had CNS dysfunction without any history or physical
find-ings suggestive of brain injury, patients who had prior known brain injury, and patients who had at least one other major malformation that was extra-cranial (Table 2). Categorizing the patients in this
way revealed that 33 individuals had developmental
delay, cerebral palsy, or mental retardation without any history or physical findings suggestive of brain injury. Of those, 12 (37%) had abnormal findings
on CT scans, including six patients in group 2 and
six patients in group 3. Four of these patients were found to have significant brain malformations char-acterized by schizencephaly, lobar
holoprosen-cephaly, white matter low-density gyral
abnormal-ities, and agenesis of the corpus callosum with a posterior fossa cyst, all unsuspected on clinical examination.
The next group consisted of 22 patients who had
CNS dysfunction secondary to a perinatal or post-natal insult. There were seven with hypoxic-is-chemic encephalopathy, six with intracranial
hem-orrhage, three with meningitis, four with spastic
cerebral palsy, one who had been physically abused, and one who had suffered a cardiorespiratory arrest. Twenty (91%) of those patients had abnormal CT scan findings including 1 1 in group 2 and nine in
group 3. Twelve of these patients also had
paren-chymal abnormalities consisting of porencephaly, infarctions, or low-density lesions. The greater like-lihood of finding an abnormality on a CT scan of the head in this group was confirmed by a
signifi-cant
x2
analysis (P = .05).The last diagnostic group included those patients who had microcephaly, plus at least one major malformation outside the CNS. Many of these
pa-tients had multiple congenital anomalies. The group consisted of 30 patients, and 20 (67%) had abnormal findings on CT scans. Of these patients, ten also had parenchymal abnormalities consisting of intracranial calcifications, primary cortical ab-normalities, migrational defects, and areas of low density. CT results were helpful in a number of instances in establishing the correct diagnosis, such as in the patients with Cockayne syndrome, intra-uterine infections, and holoprosencephaly Se-quence. Use of CT also provided us with the ability to more accurately predict the patient’s develop-mental future.
DISCUSSION
This retrospective study of 85 infants and
chil-dren who had microcephaly demonstrated that the majority of patients had abnormal CT brain scan
findings and that greater degrees of microcephaly
appeared to correlate with the presence of diffuse cerebral atrophy or ventricular dilation. This infor-mation has not been previously available in the literature.
A striking finding was the correlation of CT scan
results with the clinical category. Almost all
micro-cephalic patients who had known underlying
de-structive brain insults, such as hypoxic-ischemic encephalopathy, intraventricular hemorrhage, or
meningitis, had brain atrophy or ventricular dila-tion. In more than halfofthe patients, parenchymal disturbances were also found and were compatible
with the clinical diagnosis. Calcifications were not
present in this group.
In contrast, in patients who had CNS dysfunction
alone, with otherwise normal historical and
physi-cal findings, only slightly more than one third had
abnormal CT scan results. However, in this group, there were several instances in which the CT
find-ings were invaluable in making a diagnosis of a
previously unsuspected major brain malformation, eg, schizencephaly and holoprosencephaly, or an
intrauterine infection.
The patients who had malformations outside the
CNS were a diverse group, and the majority had
CT scan abnormalities. In several instances, the CT scan demonstrated major brain anomalies
oc-casionally found in patients with recognizable pat-terns of malformation. For example, lobar holo-prosencephaly was found in an infant of a diabetic mother,6 and schizencephaly was identified in a patient who had fetal alcohol syndrome.7 The de-gree of atrophy and distribution of intracranial calcifications in several cases helped demonstrate the degree of brain involvement and aided in the search for a specific etiologic agent.
Recognition of the specific origin of a child’s developmental disability is helpful to families for several reasons. Parents frequently have a need for
a reasonable explanation for the basis of their
child’s developmental delay and/or phenotypic ab-normalities. Having accepted this frequently elim-mates the need to search for unlikely cures and may prevent unnecessary diagnostic studies.
Estab-lishing a specific diagnosis also may provide the
child with a more definitive prognosis, particularly in the presence of brain atrophy or an underlying parenchymal lesion, because the prognosis for a child with microcephaly may be difficult to quantify at a young age.
Accurate genetic counseling is dependent on
rec-ognition of a specific diagnosis. This is not only
of malformation with an established recurrence risk in which microcephaly represents one feature, such as Cockayne syndrome or holoprosencephaly
se-quence, but in those instances in which a previously unrecognized diagnosis established by CT scan
findings leads to essentially no recurrence risk, such
as an intrauterine infection. Therefore,
micro-cephaly with or without a history of physical
find-ings suggestive of a structural anomaly is a suffi-cient indicator for a CT scan of the head. As the noninvasive techniques for evaluating brain struc-ture and function improve, new information ob-tamed using these tools could prove valuable to the patient, family, and clinician, in providing a better
understanding of the underlying disease process.
ACKNOWLEDGMENTS
This study was supported in part by Special Projects of Regional and National Significance (SPRANS) from the Public Health Service, US Department of Health and
Human Services, and the Division of Maternal and Child Health Services, Bureau for Health Services, Cabinet for
Human Resources, Commonwealth of Kentucky.
We thank Pat Jones and Beth Richardson for their
assistance in this project.
REFERENCES
1. Harwood-Nash D: Congenital craniocerebral abnormalities and computed tomography. Semin Roentgenol 1977;12:39-51
2. Ferry P: Computed cranial tomography in children. J Pe-diatr 1980;96:961-967
3. Pryor H, Thelander H: Abnormally small head size and intellect in children. J Pediatr 1968;73:593-598
4. Avery G, Meneses L, Lodge A: The clinical significance of “measurement microcephaly”. Am JDiS Child 1972;123:214-217
5. Nelhaus G: Head circumference from birth to eighteen years: Practical composite international and interracial graphs. Pediatrics 1968;41:106-114
6. Barr M Jr, Hanson JW, Currey K, et al: Holoprosencephaly in infants of diabetic mothers. J Pediatr 1983;102:565-568 7. Peiffer J, Majewski F, Fischbach H, et al: Alcohol
embryo-and fetopathy: Neuropathology of 3 children and 3 fetuses. J Neurol Sci 1979;41:125-137
A NATIONAL COLLECTION OF MATERIAL RELATED TO THE HISTORY OF
PREMATURE INFANT CARE
An effort is under way to establish a collection of materials which will trace
the development of perinatal-care technology in The National Museum of
American History at Smithsonian Institution. Readers of Pediatrics are urged
to cooperate in this national effort to locate materials of historical interest in
back rooms of hospitals and in private collections.
The material of interest ranges from incubators (Lion-type used in
incubator-baby exhibits to the present-day models), resuscitation and ventilating devices
(delivery-room apparatus, Bloxom Air-Lock, rocking bed, respirators ...),
feed-ing items (gavage equipment, nasal spoons, indwelling tubes . ..), photographs,
hospital records (statistical reports, examples of patient records ...) and
equip-ment used in landmark investigations (calorimetry, oxygen consumption ...).
Anyone who has material of interest should contact Audrey B. Davis, PhD, Curator Medical Sciences Division
The National Museum of American History Smithsonian Institution
