Acta neurol. scandinav. 57, 317-328, 1978

Department of Neurology and Neurosurgery, Medical School and Hospital de S. JoHo, Opolrto University, Oporto, Portugal

Electrophoretic pattern of cerebrospinal fluid proteins in non-neoplastic infantile hydrocephalus

CARLOS ALBERTO SILVA AND MARIA Jos6 SA

Lumbar cerebrospinal fluid (CSF) from 28 non-neoplastic hydrocephalic children was studied for total protein and electrophoretic protein pat- terns. These were classified as normal, degenerative, transudative and gamma-globulinic according to Laterre. We foiund higher total CSF pro- tein mean values than in normal cases with the same age and abnormal electrophoretic patterns in 72 YO of the cases, of which degenerative was the most common (54 %). Gamma-globulinic and transudative patterns were found in 11 % and 7 % of the cases, respectively. Several factors which may explain the increase in the total CSF protein in infantile hy- drocephalus are described: low age, ventricular block in the noncommuni- cating hydrocephalus, and probable passage of tissue proteins to the CSF from the damaged brain. The predominance of degenerative patterns may be explained by the enrichment of the CSF in tissue proteins resulting from the white matter damage provoked by the abnormal conditions of production, flow and absorption of the CSF in hydrocephalus.

Ventricular CSF was studied in four cases, and the results obtained are in agreement with the above-mentioned findings.

Clinical and experimental investigation in hydrocephalus has been mainly concerned with the cerebrospinal fluid (CSF) dynamics and etiological fac- tors. Few studies about CSF protein are found.

In normal conditions, CSF proteins increase from the ventricles to cis- terna magna, lumbar and cerebral subarachnoid spaces (Hill et al. 1959, Laterre 1965, Davson 1967, Hochwald & Wallenstein 1967, Hochwald 1970) suggesting a passage of proteins (and other substances) from nervous tissue to CSF during its circulation. This has been entitled “sink action” by Davson (1967). This clearance of the CSF, similar to that of the lym- phatic system, promotes removal of proteins (and other substances) from the brain and spinal cord (Davson 1967, Davson et al. 1970, Tourtellotte 1970, Lancet 1975, Hochwald et al. 1976).

Most cases of hydrocephalus are obstructive (RansohofJ et a2. 1960, Cutler et al. 1973) and obstruction to the CSF pathways may interfere with the clearing function and, as a consequence, with the protein composition of the CSF. Some reports have shown a rise of total protein and abnormali-

318

ties in protein electrophoretic pattern in the CSF of infantile hydrocephalus (HiZE et aE. 1959, Laterre 1965). Laterre (1965) found low total protein, increase in pre-albumin and very low gamma-globulin in the ventricular CSF when a ventricular block was present, except in some cases of infantile hydrocephalus (all younger than 2 years) where high total protein and low pre-albumin were found. Hill et al. (1959) also report increased protein in the ventricular CSF in infantile communicating hydrocephalus.

In the present paper we describe and discuss the protein patterns of the CSF in 30 cases of non-neoplastic infantile hydrocephalus and we try to explain some of the protein alterations on a physiopathological basis.

MATERIALS AND METHODS

Thirty non-neoplastic hydrocephalic children aged from 1 month to 8 years (22 of them were less than 1 year old) were studied. CSF was sampled by lumbar puncture except in two cases where, due to lumbar myelomeningocele, ventricular CSF was sampled by percutaneous ventricular tap. In two other cases with a noncommuni- cating hydrocephalus, lumbar and ventricular CSF were both studied.

The classification of hydrocephalus in communicating and noncommunicating types was done by pneumoencephalography in 24 cases.

The hydrocephalus was related in 12 cases with etiological factors, i.e. previous in- flammatory diseases or/and congenital malformations.

Protein assays

Total protein concentration was determined in the CSF by a modified Lowry’s method (Papadopoulos et al. 1959) and in the serum by the technique of Sols (1949). We consider as normal total CSF protein values below 45 mg/100 ml in children older than 3 months (Abramowicz 1969, Balfour & Gruber 1969, Harms 1975, Siemes et al. 1975). In children below 3 months of age it has not yet been possible to define a normal maximum value.

Positive-pressure-filtration device SM 16223 and SM 12136 membranfilter, P, 25 mm (Sartorius Membranfilter GM bH, Gottingen, FRG) were used for CSF coacentra- tion. The volume of the CSF used for concentration was decreasing with the rising of the total CSF protein. Therefore, 3 ml were used when total protein was lower than 45 mg/100 ml, 1.5 ml when that value reached 100 mg/100 ml and 0.5 ml far values higher than 150 mg/100 ml. A protein concentration between 3 and 4 g/100 ml was obtained. A degree of concentration about 150 times was achieved in CSF with normal protein levels.

Electrophoresis was carried out on microscope slides coated with 1.2 % (w/v) agar gel or agarose gel in barbital buffer, pH 8 6, ionic strength 0.05. Slits for samples measuring 1.0 X 0.1 cm were made with a metal mould and a volume of 10 ,ul of concentrated CSF containing 0.3-0.4 mg of proteins or 10 p l of half-diluted serum was applied. Electrophoresis was run at 150 V for approximately 90 min. A mixture of albumin, transferrin and dextran was used as a reference marker and the run was ended when the albumin fraction had migrated 2.5 cm or 4 cm when agar gel or agarose gel were used, respectively. After electrophoresis the slides were fixed in 10 o/o acetic acid in methanol, dried under filter paper in air, stained with Amidoblack and rinsed with 5 % acetic acid. The scanning was made by a Beckmann Densitometer.

319

Table 1 . Normal values f o r agarose gel electrophoresis of lumbar CSF proteins in adults

Per cent of total protein ~~

Pre- Globulins albumin Albumin a1 Q2 P r y+SaT

6.3 57.1 4.6 8.1 11.3 5.2 7.4 x (n = 20)

s.d. 1.5 4.1 0.8 1.5 1.6 0.8 1.7

-

- x - Mean values: s.d. - Standard deviation.

Clasrification of electrophoretic pattern

The classification of Laterre in protein patterns was used since it allowed us the in- dividual study of the cases. Laterre defined three pathological patterns of lumbar CSF, which are, degenerative, transudative and gamma-globulinic patterns. A degen- erative pattern is characterized by a rise of tissue microproteins and plasma proteins of low molecular weight with loiw or normal to'tal protein and is frequent in degen- erative diseases of the central nervous system (CNS). A transudative pattern is char- acterized by a rise of total protein and of high molecular weight proteins which express an alteration of the blood-brain barrier. In gamma-globulinic pattern there is a se- lective rise of y-globulin, with normal or slightly elevated proteinorachia; this pattern is frequent in subacute or chronic inflammatory diseases of the CNS. We used to classify electrophoretic patterns: a - nosrmal pattern (see Table 1); b - degenerative pattern: all or some of the following conditions - increased pre-albumin above 10 % of total protein (%), P-globulin above 15 %, 7-globulin above 7.5 %, SaT-globulin (post-y) above 1.5 %, and decreased y-globulin belorw 3 9%; c - transudative pattern: decreased pre-albumin below 3 %, increased y-globulin above 12 % and total protein above 45 mg/100 ml; d - gamma-globulinic pattern: selective rise of y-globulin above 12 %, oligoclonal y-globulin or increased y-globulin above its concentration in serum.

For statistical evaluation of the differences between the groups of patients concern- ing total CSF protein, the Mann-Whitney U-test was used (Campbell 1967).

RESULTS

Clinical and experimental data are presented in Table 2. The relation be- tween age and the total protein in lumbar CSF is pointed out in Fig. 1 and Table 3 ; obviously the total protein decreased with increasing age.

Electrophoresis of lumbar CSF protein showed in 72 % of the samples an abnormal pattern being degenerative pattern in 54 % (Table 4). The other 13 cases were distributed as follows: eight (29 %) were normal, two (7 %) transudative, and three (11 %) were gamma-globulinic patterns.

In Tables 4 and 5 etiological factors were related to the lumbar CSF electrophoretic patterns. From eight cases with inflammatory etiology, three

320

Table 2. Infantile hydrocephalus: synopsis of results concerning total CSF

CSF

Globulin Case Age Total Pre- Albumin

% P r ?’ 2aT protein albumin

% % % 5% mgJ1OOml %

1 2 3a 4 5 6 7 8 8a 9

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26a 27 28 29 29a 30

2 yrs 5 mths 1 mth 1 mth 2 mths 3 yrs 5 yrs 2 mths 2 mths 5 mths 2 mths 9 mths

10 mths 18 mths 3 mths 9 mths 7 mths 2 mths 2 mths 9 mths 8 yrs

18 mths 3 mths 9 mths 3 mths

10 mths 4 mths

16 mths 2 yrs 2 mths 2 mths 7 mths

64 39

138 32 34 13 13

139 1630

31 61 33 37 67

151 32 20 84 49 15 31 23 29 21 34 84 19 34 10

143 69 43

2.7 46.8 4.3 59.0 1.6 46.9 8.8 37.3 5.9 48.8

10.7 43.9 5.1 59.1 1.8 59.2 1.2 48.1

59.4 7.5 55.3 5.8 63.6

14.6 59.2 2.4 48.0 1.5 52.1 1.8 62.2 7.3 58.8 5.5 53.2 6.1 50.4 9.8 79.5 2.2 71.4

13.8 55.1 6.1 48.5

12.7 57.4 7.4 60.3 3.8 47.0

12.7 34.1 6.2 49.6 8.1 65.3 2.0 47.2

4.6+4.7 35.5 10.3 31.0

6.1 6.1 9.9 4.9

15.2 12.1 9.7 4.7

12.1 6.0 11.8 2.1 12.6 2.8

11.1

7.2 14.2 8.3

10.7 21.3 25.8 15.5 14.6 22.6 15.4 4.7 8.2

13.8 12.2 14.3 7.3

15.7 9.5

12.8 8.1

23.7 17.3 10.3

8.6 7.2 3.7 7.7 1.6 9.9 3.2 3.5 7.4 5.5 3.1 0.9 2.6 6.7

11.4 3.5 2.6 3.9

14.1 3.2 4.0 3.9 7.0

15.2

26.5 4.9 8.7 4.5 9 .O

15.0 1.8

18.5 32.1 7.9 3.7 4.3 3.5 3.7 7.9 3.8 0.8 1.3 1.5 0.3 2.2 0.8 3.5 1.7 3.7

13.9 4.7 9.6 2.3 3.6 4.6 2.6

(37 %) had gamma-globulinic patterns and five had normal patterns (62 %). Two cases with congenital malformation presented degenerative patterns. In 18 cases no probable etiological factor could be discovered, and 13 of them had degenerative patterns.

As far as total lumbar CSF protein is concerned, we verified that whether a previous inflammatory process was present or not, no remarkable differ- ence existed either in the number of cases with total protein values below

32 1

protein, CSF electrophoretic values, hydrocephalus type and etiological factors

Serum Etiological factors

Hydroc. Electro-

G N GO D N G D G G N N N D D T D D D D D N D D D N T V N D D V D

type Inflammation Cosng enit a1 ma1 f orm at ion

- C - -

C

NC NC NC NC NC

C C C NC NC NC NC NC C C C C C NC

C NC NC NC C

-

-

-

Meningitis

Meningitis

Meningitis Encephalitis

Syphilis Syphilis Meningitis Meningitis Meningitis

-

-

-

- -

- - -

-

- - Meningitis - - -

-

- - -

-

-

- -

-

Myelomeningocele - -

-

- -

- Encephalocele Agenesis of corpus callosum - -

- Myelomeningocele - -

a - Ventricular CSF. CSF electrophosretic pattern (see text): N - Normal, D - Degenerative, T - Transudative, G - Gamma-globulinic, Go - Oligoclonal gamma-globulinic, V - Ventric- ular. Hydrocephalus type: C - Communicating; N C - Noncommunicating.

or above 45 mg/100 ml (Table 5 ) or in the mean total CSF protein (Ta- ble 6).

In 24 patients hydrocephalus was classified as communicating or non- communicating by pneumoencephalography. Comparisons revealed (Table

322

Table 3. Total protein values of lumbar CSF in hydrocephalic children of different age

Age Total

under 3 mths 3 mths to 2 yrs above 2 yrs (n = 28) (n = 7 ) (n = 16) (n = 5)

X 77 43 26 49 Total CSF s.d. 43 33 20 36

(mg/100 ml)

-

Protein E 32-143 15-151 10-64 10-151

- x - Mean; s.d. - Standard deviation; E - Range of extreme values. Difference between the sample medians was tested for significance using the Mann- Whitney U-test: P < 0.05 comparing the groups under 3 months and above 2 years.

5 ) that 10 (83 %) of 12 communicating hydrocephalus had total protein below 45 mg/100 ml, while only six (50 %) of 12 noncommunicating hy- drocephalus presented total protein below that value. The total protein in the noncommunicating type (mean 57 mg/100 ml) was higher than in the communicating type (mean 45 mg/100 ml) but with no statistic significance (Table 6). In both types degenerative patterns predominated (Table 5).

Ventricular CSF protein was studied alone in two cases (3a and 26a) and together with lumbar CSF in two cases more (8a and 29a). A lumbar myelo- meningocele was present in the first two cases. Case 3a complicated by a meningitis, presented elevated total CSF protein and an oligoclonal gamma-

; ] 0 200

0 .-

E l e c t r o p h o r e t i c P a t t e r n

0 n o r m a l

0 d e g e n e r a t ~ v e

A t r a n s u d a t l v e

I-globulinic

I I " ' I I I I I l l 6 12 2 3 4 5 6 7 0

m o n t h s years age

Figure 1 . Infantile hydrocephalus: relationship between age, total protein and electrophoretic pattern of lumbar CSF,

323

Table 4 . Relationship between etiological factors and electrophoretic pattern of lumbar CSF in infantile hydrocephalus

~

Number CSF electro'phoretic pattern Etiology of

cases N D T G

Inflammatiosn 8 5 (Meningitis) (6) (5) (Encephalitis) (1) (Syphilis) (1)

Congenital Malformatioln 2 2 Unknown 18 3 13 2

Total (%I

2 3 ( 7 ) (11)

CSF pattern (see text): N - Normal; D - Degenerative; T - Transudative; G - Gamma- globulinic.

globulinic electrophoretic pattern was traced. Case 26a with a myelomenin- gocele without complication had a low total CSF protein and a ventricular CSF pattern. Ventricular and lumbar CSF proteins were compared in cases 8 and 29 (noncommunicating hydrocephalus). Congenital syphilis (case 8) gave rise to a high total protein and a gamma-globulinic electrophoretic pattern both in ventricular and lumbar CSF. Opposite to these findings, case 29 showed elevated total protein, an high and split pre-albumin and an increase of z and 6aT globulins in ventricular CSF, while in lumbar CSF increased total protein and a degenerative electrophoretic pattern, in spite of the elevated total protein and decreased pre-albumin, were present.

DISCUSSION

In the present paper we used the classification of electrophoretic protein patterns of Laterre (1965). We think it is valid to use that classification, based on values of adult lumbar CSF, for the CSF of children, as far as the differences usually found between those two age groups are considered. According to that classification and to the results of Siemes et al. (1975) the usual electrophoretic pattern of the newborn until 2 weeks is similar to the transudative pattern of adults. From 2 weeks till 3 months of age, transudative patterns are still expected to appear. In children between 3 months and 2 years of age, Siemes et al. (1975) describe a rise of pre- albumin and ,.%globulin which is not enough to reach the pathological values of degenerative pattern of Laterre.

Tab

le 5

. Tot

al p

rote

in v

alue

s an

d el

ectr

opho

reti

c pa

tter

n of

lum

bar

CSF

con

side

ring

in

flam

mat

ory

etio

logy

and

hyd

roce

phal

us t

ype

Tot

al C

SF p

rote

in (

mgI

100

ml)

C

SF e

lect

roph

oret

ic p

atte

rn

<45

>45

N

D

T

G

Infl

amm

atio

n 3

-

- Pr

esen

t n

=8

5

3 n

=8

5

(%)

(100

) (6

2)

(37)

(1

00)

(62)

(3

7)

Abs

ent

or

n =

20

14

6 n

= 2

0 3

15

2 un

know

n (%

) (1

00)

(70)

(3

0)

( 100

) (1

5)

(75)

(1

0)

-

Tot

al

(%)

n =

28

19

9 n

= 2

8 8

15

2 3

(100

) (6

8)

(32)

(1

00)

(29)

(5

4)

(7)

(11)

Tot

al

(%I

~~~~

~~

~ ~

CSF

ele

ctro

phor

etic

pat

tern

(se

e te

xt):

N -

Nor

mal

; D

- D

egen

erat

ive;

T - T

rans

udat

ive;

G - G

amm

a-gl

obul

inic

.

325

Table 6 . Total protein values of lumbar CSF considering inflammatory etiology and hydrocephalus type

Inflammatio'n Hydrocephalus type

Present Absent or Communi- Non-

communi- cating

(n = 12)

unknown cating (n = 20)

(n = 8) (n = 12)

X 51 47 45 57 Total CSF s.d. 37 39 34 45 Protein E 13-138 10-151 21-151 10-143 (mg/100 ml)

i

X - Mean; s.d. - Standard deviation; E - Range of extreme values. Difference of medians was not significant using the Mann-Whitney U-test.

Our findings are in agreement with the results obtained by Hill et al. (1959) and Laterre (1965) who studied a few cases of infantile hydro- cephalus. It is suggested that the age alone gives no support to an increase in total CSF protein and, therefore, other factors may be responsible.

Electrophoresis of CSF proteins revealed a great predominance of ab- normal patterns especially of degenerative pattern (54 96) (Table 4).

Our data revealed that the etiology of hydrocephalus may influence the electrophoretic patterns. In general, previous inflammatory diseases did not influence the total CSF protein, but they may explain the presence of gam- ma-globulinic patterns (Laterre 1965).

Another factor which may influence total protein and electrophoretic pat- terns is the presence or absence of a ventricular block, that is, whether hy- drocephalus is of communicating or noncommunicating type. The increase of proteins and decrease of pre-albumin in lumbar CSF described by Hill et al. (1959) when a ventricular block exists is partially in agreement with our results. However, degenerative electrophoretic patterns predominated in both types of hydrocephalus, in spite of increase in total protein.

Either in communicating or in noncommunicating hydrocephalus, CSF is prevented from reaching its usual place of absorption, the arachnoid villi. As the formation of CSF either by the choroid plexuses or by the ventricular ependyma continues at the same rate according to some authors (Bering & Sat0 1963, Cutler et al. 1968, Lorenzo et al. 1970) or at a lower rate according to other authors (Sahar et al. 1970, Hochwald et al. 1972), an alternate pathway of absorption may work. A transventricular pathway with passage of the CSF to the subependymary white matter and from here to the neighbouring venous vessels may be operative (Bering & Sat0 1963, Sahar et al. 1969, Milhorat et al. 1970a, Strecker et al. 1974, Price et al. 1976).

326

Alternatively, absorption through the choroid plexuses (Milhorat et 1970b, Dohrmann 1971, Lawson & Raimondi 1973) and even a spi route of absorption in communicating hydrocephalus may work (Dav, et al. 1970, DiChiro et al. 1976). The continued production of the CSF a space with blocked exit provokes ventricular dilatation with enlargemc and structural alterations of the ventricular surface and the periventricu white matter (Weller & Wi$niewski 1969, Clark & Milhorat 1970, Lux et 1970, Milhorat et al. 1970a, Sahar et al. 1970, Hochwald et al. 1972, Well & Shulman 1972, Lawson & Raimondi 1973, Price et al. 1976). Disco tinuity of the ependymary epithelium and functional changes of ependyn increase CSF absorption by that transependymary route (Page 1975). Vei tricular dilatation is mainly due to the loss of proteins and lipids from th damaged white matter provoked by the rise in the intraventricular pressur and transependymary flow of the ventricular CSF (Fishman & Greer 1963 Weller & Wihiewski 1969, Milhorat et al. 1970a, Weller & Shulman 1972 Weller & Williams 1975). Liberated proteins may go to the ventricular CSI (Hochwald et al. 1972, Lawson & Raimondi 1973) through a stretched ana torn ependyma in which disruption of intercellular junction is present (Wel- lev & Wis’niewski 1969, Milhorat et al. 1970a, Lawson & Raimondi 1973, Price et al. 1976), seeing that, after some time, an equilibrium between for- mation and absorption of CSF is reached and intraventricular pressure re- turns to lower and even normal values (Lorenzo et al. 1970, Lux et al. 1970, Hochwald et al. 1972).

This physiopathological mechanism may then explain the enrichment in tissue proteins of the CSF in hydrocephalus.

The report of Forslund et al. (1976) showing a great rise of total protein in lumbar CSF from shunt-treated hydrocephalic children, lasting for several vears after shunt operation, confirms the influence of an abnormal path- way of circulation and absorption of the CSF in the rise of proteins in lumbar CSF, probably through a reduction of the spinal’flow provoked by the shunted CSF circulation.

ACKNOWLEDGEMENTS

We are grateful to the members of the Neurology Department Laboratory €or tech- nical assistance and to the medical staff of the Neurology and Neurosurgery Depart- ment for clinical evaluation of hydrocephalic children. We are also indebted to Pro- fessor Joaquim Maia for advice on statistics.

REFERENCES

Abramowicz, M. (1969): Normal values for cerebrospinal fluid protein concentration in

Balfour, H. H. & Gruber, M. L. (1969): On redefining the normal cerebrospinal fluid children. Clin. Pediat. 8, 300-301.

327

protein concentration in children. Clin. Pediat. 8, 22A-23A. Bering, E. A. & Sato, 0. (1963): Hydrocephalus: changes in formation and absorption

of cerebrospinal fluid within the cerebral ventricles. J . Neurmurg. 20, 1050-1063. Campbell, R. C. (1967): Statistics for biologists, pp. 44-46. Cambridge University Press,

Cambridge. Clark, R. G. & Milhorat, T. H. (1970): Experimental hydrocephalus. Part 3: light mi-

croscopic findings in acute and subacute obstructive hydrocephalus in the monkey. J. Neurosurg. 32, 400-413.

Cutler, R. W. P., Page, L., Galicich, J. & Watters, G. V. (1968): Formation and absorp- tion of cerebrospinal fluid in man. Brain 91, 707-720.

Cutler, R. W. P., Murray, J. E. & Moody, R. A. (1973): Overproduction of cerebrospinal fluid in communicating hydrocephalus. Neurology 23, 1-5.

Davson, H. (1967): Physiology of the Cerebrospinal Fluid. Churchill, London. Davson, H., Hollingsworth, G. & Segal, M. B. (1970): The mechanism of drainage of

the cerebrospinal fluid. Brain 93, 665-678. DiChiro, G., Hammock, M. K. & Bleyer, W. A. (1976): Spinal descent of cerebrospinal

fluid in man. Neurology 26, 1-8. Dohrmann, G. J. (1971): The choroid plexus in experimental hydrocephalus. A light

and electron microscopic study in no+mal, hydrocephalic, and shunted hydro- cephalic dogs. J. Neurosurg. 34, 56-69.

Fishman, R. A. & Greer, M. (1963): Experimental obstructive hydrocephalus. Changes in the cerebrum. Arch. Neurol. 8, 156-161.

Farslund, M., Bjerre, I. & Jeppson, J. 0. (1976): Cerebrospinal-fluid protein in shunt- treated hydrocephalic children. Dev. Med. Child. Neurol. 18, 784-790.

Harms, D. (1975): Comparative quantitation of immunoglobulin G (IgG) in cerebro- spinal fluid and serum of children. Europ. Neurol. 13, 54-64.

Hill, N. C., Goldstein, N. P., McKenzie, B. F., McGuckin, W. F. & Svien, H. J . (1959): Cerebrospinal-fluid proteins, glycoproteins, and lipoproteins in obstructive lesions of the central nervous system. Brain 82, 581-593.

Hochwald, G. M. & Wallenstein, M. (1967): Exchange of albumin between blood, cere- brospinal fluid, and brain in the cat. Am. J. Physiol. 212, 1199-1204.

Hochwald, G. M. (1970): Influx of serum proteins and their concentrations in spinal fluid along the neuroaxis. J. neurol. Sci. 10, 269-278.

Hochwald, G. M., Lux, W. E., Sahar, A. #& Ransohoff, J. (1972): Experimental hydro- cephalus. Changes in the cerebrospinal fluid dynamics as a function of time. Arch. Neurol. 26, 120-129.

Hochwald, G. M., Wald, A. & Malhan, C. (1976): The sink action of cerebrospinal fluid volume flow. Effect an brain water content. Arch. Neurol. 33, 339-344.

Lancet (1975): Cerebrospinal fluid: the lymph of brain? Lancet ii, 444-445. Laterre, E. C. (1965): Les ProtCines du Liquide CCphalo-rachidien B l’etat Normal et

Pathologique. Arscia, Bruxelles. Lawson, R. F. & Raimondi, A. J. (1973): Hydrocephalus-3, a murine mutant: I. Al-

terations in fine structure of choroid plexus and ependyma. Surg. Neurol. l, 115- 128.

Lorenzo, A. V., Page, L. K. & Watters, G. V. (1970): Relationship between cerebro- spinal fluid formation, absorption and pressure in human hydrocephalus. Brain 93,

Lux, W. E., Hochwald, G. M., Sahar, A. & Ransohoff, J. (1970): Periventricular water content. Effect of pressure in experimental chronic hydrocephalus. Arch. Neurol. 23, 475-479.

679-692.

328

Milhorat, T. H., Clark, R. G., Hammock, M. K. & hlcGrath, P. P. (1970a): Structural, ultrastructural, and permeability changes in the ependyma and surrounding brain favoring equilibration in progressive hydrocephalus. Arch. Neurol. 22, 397-407.

Milhorat, T. H., Mosher, M. B., Hammock, M. K. & Murphy, C. F. (1970b): Evidence for choroid-plexus absorption in hydrocephalus. New Engl. J. Med. 283, 286-289.

Page, R. B. (1975): Scanning electron microscopy of the ventricular system in normal and hydrocephalic rabbits. Preliminary report and atlas. J. Neurosurg. 42, 646-664.

Papadopoulos, N. M., Hess, W. C., O’Doherty, D. & McLane, D. E. (1959): A procedure for the determination of cerebrospinal fluid total protein and gamma-globulin in neurologic disorders. Clin. Chem. 5, 569-574.

Price, D. L., James, A. E., Sperber, E. & Strecker, E.-P. (1976): Communicating hydro- cephalus. Cisternographic and neuropathologic studies. Arch. Neurol. 33, 15-20.

Ransohoff, J., Shulman, K. & Fishman, R. A. (1960): Hydrocephalus. A review of etiol- ogy and treatment. J. Pediat. 56, 399-411.

Sahar, A., Hochwald, G. M. & Ransohoff, J. (1969): Alternate pathway for cerebro- spinal fluid absorption in animals with experimental obstructive hydrocephalus. Exp. Neurol. 25, 200-206.

Sahar, A , Hochwald, G. M. & Ransohoff, J . (1970): Experimental hydrocephalus: cere- brospinal fluid formation and ventricular size as a function of intraventricular pressure. J. neurol. Sci. 11, 81-91.

Siemes, H., Siegert, M. & Rating, D. (1975): Das Liquorproteinprofil normaler Kinder und seine Abhangigkeit vom Lebensalter. Untersuchungen mittels CAF- und Agarosegel-Elektrophomrese. Neuropadiatrie 6 , 383-397.

Sols, A. (1949): Valoration de las proteinas del suer0 por referencia a la media normal. Rev. Esp. Fisiol. 5, 239-250.

Strecker, E.-P., James, A. E., Konigsmark, B. & Merz, T. (1974): Autoradiographic ob- servations in experimental communicating hydrocephalus. Neurology 24, 192-197.

Tourtellotte, W. (1970): On cerebrospinal fluid immunoglobulin-G (IgG) quotients in multiple sclerosis and other diseases. A review and a new formula to estimate the amount of IgG synthesized per day by the central nervous system. J. neurol. Sci.

Weller, R. 0. & WiSniewski, H. (1969): Histological and ultrastructural changes with

Weller, R. 0. & Shulman, I(. (1972): Infantile hydrocephalus: clinical, histological, and

Weller, R. 0. & Williams, B. N. (1975): Cerebral biopsy and assessment of brain dam-

10, 279-304.

experimental hydrocephalus in adult rabbits. Brain 92, 819-828.

ultrastructural study of brain damage. J. Neurosurg. 36, 255-265.

age in hydrocephalus. Arch. Dis. Child. 50, 763-768.

Received June 3, 1977, accepted February 27, 1978

Carlos Albert0 Silva, M.D. Serviso de Neurologia e Neurocirurgia Faculdade de Medicina Porto Portugal