CT evaluation of normal CSF spaces in children: relationship to age, gender and cranial size

EUROPEAN JOURNAL OF RADIOLOGY

ELSEVIER SCIENCE IRELAND

European

Journal

of Radiology

18 (1994) 22-25

CT evaluation of normal CSF spaces in children: relationship to age, gender and cranial size P. Prassopoulosa, “Department hDeparrmenr

ofMedical

of Diagnosric

Instrumenls.

Radiology.

(Received

University

School of Technological Street.

22 June

1993; revision

D. Cavouras* b Hospiral.

Applications,

Kallirhea

received

Medical

School q/ Crere, Crete. Greece

Technologicul

17671, Athens,

10 September

Educational

Institute

of Athens,

37-39

Esperidon

Greece

1993; accepted

29 September

1993)

Abstract The extent of the cerebrospinal fluid (CSF) spaces was measured in 247 CT examinations, reported as normal, in children aged 3 months to 14 years. The measurements of the CSF compartments were divided by the sum of the transverse and longitudinal internal cranial diameters in the corresponding CT section, in order to take into account the size and shape of the growing skull. All CSF spaces were relatively larger in the younger (I 3 years) than in the older children, but did not differ between boys and girls, since the cranial size was taken into account. The CSF compartments increased in a non-uniform manner during the first 3 years of life, but after the age of four they develqped uniformly and in parallel with the growing skull. Measurements of the subarachnoid spaces formed an age-related table, which may be of value when interpreting brain CT examinations of children. Key

words:

Brain, CT; Brain, infants and children; Computer tomography, brain: Children. central nervous system

1. Introduction Several studies have shown that in children

undergo-

ing chemotherapy [l-3], treated with corticosteroids [4], or with a history of cranial trauma [S], the subarachnoid space may be enlarged. On the other hand, in children with post-traumatic cerebral edema [5] or in non-traumatic coma [6] the cerebrospinal fluid (CSF) space may be reduced. Also, the extent of various CSF compartments changes in case of hydrocephalus [7] or in benign processes, such as in children with widening of the frontal subarachnoid space [8]. However, in assessing these changes on CT, it is necessary to have knowledge of the normal extent of the subarachnoid space during childhood. Previous studies have evaluated the normal CSF spaces in children by visual inspection [9] or by linear measurements of selected supratentorial CSF compartments [ 10,111. However, the subjective evaluation of the CSF spaces cannot be accurately reproduced and linear measurements of the width of the * Corresponding 0720-048X/94/$07.00 SSDI

author.

0

0720-048X(93)00465-8

1994 Elsevier

Science Ireland

CSF compartments do not take into account the variation in cranial size and shape among children. In this study the main supra- and infratentorial normal CSF spaces were measured by CT and were related to the child’s age, gender and cranial size. The aim was to provide data for an accurate and reproducible evaluation of the relative extent of the CSF compartments in the growing skull. 2. Materials and methods This study comprised 247 brain CT examinations, reported as normal, of 145 boys and 102 girls aged 3 months to 14 years. Patients with cerebral focal lesion, evidence of increased intracranial pressure, abnormal head circumference, central nervous system infection, chronic use of steroids or chemotherapeutic agents, history of perinatal asphyxia or any other cause that may alter the morphology of the brain or skull were not included in this study. Patients underwent CT for simple head trauma, sporadic convulsions, headache, abnormal EEG or staging of extracranial malignant tumor.

Ltd. All rights reserved

P. Prassopoulos. D. Cuvourav/ Eur. J. Radiol. 18 (1994) 22-25

23

b

Fig. I. Selected measurements of (a) the lateral ventricles (LV). (b) the anterior horns (AH) and the bicaudate nuclei distance (BN). (c) the babal cistern (BC). and (d) the prepontine cistern (PC) and the fourth ventricle (FV). All measurements were divided by the sum of the respective transverse (TT) and longitudinal (LL, skull diameters.

24

P. Prussopoulos,

All patients were examined on a Phillips LX CT scanner, with 5-mm contiguous CT sections in infants and 10 mm in older children, except for the posterior fossa, where 5-mm sections were used. Employing the CT software the following measurements (Fig. 1) were performed: the minimum width of the bodies of the lateral ventricles (LV), the maximum distance between the anterior horns (AH), the minimum bicaudate nuclei (BN) distance, the maximum width of the third (TV) and fourth (FV) ventricles, the anteroposterior diameter of the prepontine cistern (PC), the two perpendicular diameters of the basal cistern (BC), and the area enclosed by the maximum internal perimeter of the skull at the midventricular level. Each CSF measurement was divided by the sum of the transverse (TT) and the longitudinal (LL) internal skull diameters, measured at the corresponding CT section (Fig. l), in order to take into account the cranial size and shape. The so formed ratios (CSF indices) were used to evaluate the variation of the CSF compartments in relation to the size of the growing skull. Statistical analysis was performed on a computer employing Student’s t-test and the parametric regression line analysis. 3. Results The values of the normal CSF indices during childhood a.re presented in the age-related Table 1. All CSF indices were significantly higher in the younger (I 3 years) than in the older children (PLvI < 0.025, PAH, < 0.01, Pam < 0.001, Prv, c 0.05, Prvr < 0.05, Pat, < 0.01, Ppcl < 0.05). No significant differences in

Table I Normal values of seven CSF indices ( x IO--‘) and the sum of maximal Years

Patients

BNI

AHI

LVI

TV1

D. Cuvouros/

Eur. J. Rudiol.

18 (1994)

22-25

CSF indices were found between boys and girls, either younger or older than 3 years. In children aged 4-14 years none of the CSF indices varied with age (0.051 < r c 0.491, P > 0.10). However, in the younger children ( 5: 3 years), age-related differences in the CSF indices were observed: the lateral ventricles were larger in the second than in the first year of life (PLvI, PAHI, PaNi < 0.025) and the prepontine and basal cisterns were larger in the first than in the second year of life (PpcI, Paci < 0.025). The third and the fourth ventricles did not differ significantly between the first and the second or the second and the third years of life (P > 0.10). Strong correlation existed between the sum of the maximum longitudinal and transverse internal skull diameters and the maximum area enclosed by the internal perimeter of the skull at the midventricular level (r = 0.948, P < 0.001). That correlation was higher than the correlation between the maximum transverse diameter taken alone and the maximum skull area (r = 0.595, P < 0.05). 4. Discussion The volume of the subarachnoid space may be affected by a variety of conditions [1,2,5,6,12]. Abnormal changes in the size of the CSF space in children have been evaluated mainly by visual inspection on CT. A more accurate and reproducible assessment of local or diffuse changes of the subarachnoid spaces may be achieved by measuring the extent of the CSF compartments. This quantitative evaluation requires the

skull diameters (cm) during childhood FVI

BCI

TT+LL

PCI

Boys

Girls

O-l

I5

106 f

12

108 *

I2

34 l 5

8zt2

78 f

9

I48 zt I2

47 f

8

25.8 f

1.3

23.4 f

I.1

l-2

26

II5

II

II8

f

I3

38 f

5

9*3

76 f

II

139 f

40*

10

27.3 f

1.4

26.1 f

I.5

2-3

21

102 l I4

II2

f

II

34 f

6

8zt3

79 f

8

I31 l I4

38 f

II

27.6 +z 0.8

26.8 f

I.9

3-4

20

103 f

I2

104 f

II

29 + 6

8zt2

74 f

8

I28 l 7

38 f

IO

28.2 f

I.1

27.2 f

0.8 I.1

f

II

4-5

29

99*

I3

106 f

7

28 f

4

7*2

12 f

9

130 iz 8

42 f

I2

28.4 zt I.4

27.2 f

5-6

22

103 f

I2

108 f

8

30 f

6

7*3

75 f

7

124 f

II

40 l II

28.4 l 1.9

27.4 l 2.1

6-7

15

104 f

13

109 t

I2

29 l 5

7*l

76 l 7

126 l 9

38 zt 9

28.4 f

2.3

27.5 f

1.7

7-8

I3

I05 l II

IO8 f

II

28 f

5

8a2

75 f

9

122 f

35 f

8

28.2 f

3.1

27.4 f

2.6

I05 l 9

31 f

5

9*3

72 f

II

130 l 9

31

9

28.3 f

I.1

27.8 f

I.3

107 f

28 zt 5

8*2

76 zt 8

I25 f 9

37 f

IO

28.7 f

I.2

27.8 & I.2

7*1

75 l 8

125 + 10

35 f 9

29.0 + I.7

27.6 f

1.6

8&l

75 f

9

130 f

IO

33 f

8

29.1 f 0.8

28.0 f

I.9

9

39 f

9

29.7 f

1.8

28.3 f

1.4

8

l

8-9

I4

98 f

9-10

12

98 + I3

IO-11

I3

103 + 13

I07 l II

29 f

II-12

II

I05 f

I3

105 f

8

31 *7

12-13

IO

101 f

I5

103 f

7

32 f

6

7zt2

17 f

I2

134 f

13-14

I4

102 f

I4

105 f

7

28 f

5

7*l

78 f

6

134 l 7

32 l 7

29.8 f

2.1

28.4 f

I.8

14-15

I2

I01 f

15

106 f

8

31 *6

7*2

13 f

9

130 f

36 f

29.5 f

2.2

28.3 f

I.6

LVI,

14

lateral ventricles index; AHI,

5

anterior horns index: BNI, bicaudate nuclei index; TVI,

basal cistern index; PCI, prepontine of patients per year group.

9

cistern index; TT + LL, sum of the maximum

II

7

third ventricle index: FVI.

transverse and longitudinal

fourth ventricle index; BCI.

skull diameters: Patients. number

P. Prassopoufos. D. Cavouras/ Eur. J. Radiol. 1% (f!W4) 22-Z

establishment of normal data of the CSF spaces during childhood. Although in adults such data have been extensively reported [13], in children there are few reports on this subject [lO,l I]. In the latter studies, the width of various supratentorial CSF compartments was measured on Polaroid films and no attention was given to the influence that the cranial size and shape may have on the extent of the subarachnoid space. However, it is known [ 14,151 that the size of the skull varies among normal children of the same age and it is also age- and gender-dependent. Therefore, in the present study all measurements of the CSF compartments were referred to the cranial size in order to examine the relative extent of the normal subarachnoid spaces during childhood. According to our data the ratio of the transverse to the longitudinal maximal skull diameters varied between 0.64 and 0.93, indicating significant skull shape variations among children. It was also found that a high correlation existed between the sum of these two diameters and the area within the maximal internal skull perimeter, which reflects the cranial size [15]. Consequently, the CSF indices of this study were formed by dividing each CSF space measurement by the sum of the transverse and longitudinal internal skull diameters and not by the transverse diameter alone, as has been done previously in adults [ 131. Employing these CSF indices, the influence that the child’s age or gender may have on the subarachnoid spaces was examined, irrespective of constitutional variations in the size and shape of the skull. Four indices previously used in adults [13] were accordingly modified (LVI, AHI, BNI. TVI) and three new indices (PCI, BCI. FVI) were introduced, since there are no studies on the fourth ventricle and the prepontine and basal cisterns in children. The cortical sulci, the Silvian and interhemispheric fissures, and the space between the cerebral cortex and the cranial vault were not measured because they are small and liable to partial volume effect errors, especially in older children [9]. However, these CSF compartments differed in size among younger children (5 3 years) from barely visible to definitely evident. This is in agreement with a previous report [9] that there is variability in the size of the subarachnoid spaces during the first 2 years of life. Also, the superior cerebral cistern and cisterna magna, although measurable, were not evaluated because they varied considerably in size among normal children of the same age. The results of this study indicated that in children younger than 3 years the various CSF compartments do not grow uniformly; the lateral ventricles are relatively larger in the second year, the basal and prepontine cisterns in the first year, but the third and fourth ventricles remain constant during the first 3 years of life. In contrast, in children aged 4-14 years the CSF com-

25

partments develop in a uniform manner and in parallel with the growing head. According to the findings of this work the extent of the subarachnoid spaces, in relation to the cranial size, is larger in the younger (13 years) than in the older children and it is not genderdependent. Differences in the CSF volume between younger and older children have been reported by other authors [9,11]. The proposed CSF indices can be of value in the interpretation of the subarachnoid spaces in children undergoing CT. Furthermore, the degree of brain atrophy induced by diseases or therapeutic agents might be systematically evaluated by employing these indices. 5. References 1

2

3

4

5 6

Enzmann DR. Lane B. Enlargement of subarachnoid spaces and lateral ventricles in pediatric patients undergoing chemotherapy. J Pediatr 1978; 92: 535-539. Kretzschmar K. Gutjahr P. Kutzner J. CT studies before and after CNS treatment for acute lymphocytic leukemia and malignant non-Hodgkins lymphoma in childhood. Neuroradiology 1980: 20: 173-180. Browers P. Poplack D. Memory and learning sequelae in longterm survivors of acute lymphoblastic leukemia: association with attention deficits. Am J Pediatr Hematol Oncol. 1990; 12: 174-181. Lagenstein tomography amethasone:

1. Willig RP, Kuhne D. Cranial computed findings in children treated with ACTH and dexfirst results. Neuropediatrics 1979; IO: 370-375.

Barkovich AJ. Metabolic and destructive brain disorders. In: Pediatric neuroimaging. New York: Raven Press, 1990; 35-77. Tasker RC, Matthew DJ, Kendall B. Computed tomography in the assessment of raised intracranial pressure in non-traumatic

8

coma. Neuropediatrics 1990: 2 I : 9 I-94. Barkovich AJ. Hydrocephalus. In: Pediatric York: Raven Press. 1990: 205-226. Odita JC. The widened frontal subarachnoid

9

parative study between the macrocephalic. microcephalic. and normocephalic infants and children. Child Nerv Syst 1992; 8: 36-39. Kleinman PK. Zito JL. Davidson RI. Raptopoulos V. The

I

IO

II

I2

13 14 I5

neuroimaging.

New

space: a CT com-

subarachnoid spaces in children: normal variations in size. Radiology 1983; 147: 455-457. Fukuyama Y. Miyao M, lshizu T. Maruyama H. Developmental changes in normal cranial measurements by computed tomography. Child Neurol 1979; 21: 425-432. Pedersen H. Gyldensted M. Gyldensted C. Measurements of the normal ventricular system and supratentorial subarachnoid space in children with computed tomography. Neuroradiology 1979: 17: 231-237. Ito M. Takao T, Okuno T. Mikawa H. Sequential CT studies of 24 children with infantile spasms on ACTH therapy. Dev Med Child Neural 1983; 25: 475-480. LeMay M. Radiologic changes of the aging brain and skull. AJR 1984; 143: 383-386. Gordon IRS. Measurement of cranial capacity in children. Br J Radio] 1966; 39: 377-382. Hahn FJ. Chu WK. Cheung JV. CT measurements growth: normal subJects. AJR 1984; 142: 1253-1255.

of cranial