Acta neurol. scandinav. 63, 255-266, 1981

Department of Neurology and Neurosurgery, School of Medicine, University of Oporto, and Hospital de S. JoCo, Oporto, Portugal

Protein patterns of the cerebrospinal fluid of 30 patients with Subacute Sclerosing Panencephalitis (SSPE)

CARLOS ALBERTO SILVA, MARIA EDITE RIO AND CELW CRUZ

Lumbar cerebrospinal fluid (CSF) from 30 cases of subacute sclerosing panencephalitis (SSPE) was studied for total protein and protein electro- phoresis. Total CSF protein values were normal in almost all the cases, but the 7-globulin was increased and had an oligoclonal morphology in all the cases. Absolute and relative values of some of the other electro- phoretic fractions were decreased. Total protein and electrophoretic se- rum values in 25 cases did not differ significantly from the controls, al- though the y-globulin presented an oligoclonal morphology in 13 cases. CSF immunoglobulins were determined in 10 cases; IgG was increased in all; IgA was traceable and slightly increased in only four cases; IgM was not found. These findings point to the intrathecal synthesis of IgG and the absence of barrier impairment. The decrease of the absolute values of some of the other CSF proteins suggests that homeostatic mechanisms for protein concentration in the CSF are maintained in SSPE.

Key words: CSF protein - CSF electrophoresis - oligoclonal y-globulin - CSF immunoglobulins - CSF protein homeostasis - SSPE.

Subacute sclerosing panencephalitis (SSPE) is a slow virus infection of the central nervous system (CNS) (ter Meulen & Hall 1978) in which very high values of y-globulin are found in the cerebrospinal fluid (CSF) as a conse- quence of the synthesis of antibodies against a SSPE virus by mononuclear cells infiltrating the CNS (Cutler et al. 1968, Link et al. 1973, Vandvik & Norrby 1973, Norrby et al. 1974), itself a probable genomic variation of the measles virus (Raine & Fields 1974).

CSF protein electrophoresis in SSPE as well as the quantitation of im- munoglobulins has, in addition to its diagnostic value, an important role in the elucidation of immunological mechanisms of the CNS and has been the subject of a detailed study (Cutler et al. 1967, 1970, L i m o et al. 1971, Link et al. 1973, Vandvik & Norrby 1973, Vandvik & Skrede 1973, Norrby et al. 1974, Peter et al. 1974, Ebers et al. 1979).

The purpose of this paper was to discuss the protein pattern of the CSF from 30 cases of SSPE observed in our department over the past 10 years.

0001-6314/81/040255-12 $02.50/0 @ 1981 Munksgaard, Copenhagen

256

MATERIAL AND METHODS

Patients. 30 patients whose ages ranged from 3 to 32 years had a confirmed diagnosis of SSPE by at least three of the following parameters: clinical, electroencephalographic, CSF protein electrophoretic, CSF serological and neuropathological typical features; 28 (93 %) of them had a typical course of the disease; 27 (90 %) had characteristic Radermecker complexes on the electroencephalogram; all 30 patients had oligoclonal y-globulinic pattern in CSF protein electrophoresis; the CSF measles antibody titers determined in 21 patients were increased in all the cases; and typical ultrastructural features were found in all the 14 cases studied. Twenty-six of the patients were in stage I1 (Zeman & Kolar 1968) and four in stage I11 of the disease at the time of the CSF sampling for protein studies.

Laboratory methods. Total CSF protein concentration was determined by a modified Lowry's method (Papadopoulos et al. 1959) and serum protein concentration by the technique of Sols (1949). Agar or agarose-gel electro- phoresis of serum and concentrated CSF proteins was carried out as has been previously described (Silva & Sa' 1978). Evaluation of the electro- phoretic fractions stained with amidoblack was made by densitometry at 580 nm wave-length (Cellomatic-2, Chemetron, Italy). Normal total protein and relative electrophoretic fraction values for serum and CSF are given in Table 1. Absolute values of the CSF electrophoretic fractions were esti- mated from the total protein and the relative electrophoretic values (Liano rt al. 1971) and are indicated in Table 2.

Immunoglobulin G (IgG), A (IgA) and M (IgM) were determined in diluted serum and unconcentrated CSF by a modification of the single radial im- munodiffusion method described by Mancini & Heremans (1965). Com- mercial monospecific rabbit antisera against human serum IgG, IgA and IgM were used (Behringwerke, FRG). The lower limits for detection were as follows: IgG, 0.15 mg/100 ml; IgA, 0.12 mg/100 ml; IgM, 0.20 mg/ 100 ml. The upper normal value for CSF IgG was estimated to be 10 % of total protein, that is, 4.5 mg/100 ml. The upper normal value for CSF IgA was not determined as this only reached the lower detection limit. CSF IgM was not detectable. Serum average values obtained from blood donors were as follows, expressed by the median and the 10 and 90 per- centile values: IgG, 1413 mg/100 ml (1044-2017); IgA, 263 mg/100 ml (125-544); IgM, 126 mg/l00 ml (69-191).

Statistical methods. The Mann-Whitney U-test (Campbell 1967) was used for the statistical evaluation of differences in total CSF and serum protein and electrophoretic values between SSPE and controls.

257

The correlation between total CSF protein and y-globulin in SSPE and controls was studied using the Pearson coefficient of correlation and the Sewal-Wright path coefficient (Li 1955).

RESULTS

In the 30 cases of SSPE studied, the median value of total CSF protein was significantly increased in relation to the median value in controls (Table 1). Only five cases (6.7 %), had protein concentrations over 50 mg/ 100 ml (Fig. 1).

The most important abnormality observed in CSF protein electrophoresis was the increased y-globulin, whose relative and absolute median values were five (Table 1) and six (Table 2) times greater than normal. The other CSF electrophoretic fractions were decreased, particularly the relative values of albumin, a2 and /?-globulin (Table 1) and the absolute values of albumin and /?-globulin (Table 2). CSF y-globulin percentage was plotted against the corresponding total

protein value (Fig. 1). Increased y-globulin and normal total protein was observed in 20 cases (66.7 %). Two cases showed normal relative values of y-globulin and total protein, and eight cases showed sligthly raised total protein and high relative values of y-globulin. There was a significant cor- relation (r = 0.4605, P < 0.02) between the y-globulin (as expressed by relative values) and the total protein concentration in the CSF of SSPE patients (Fig. 1). On the contrary, the coefficient of correlation in normal CSF was not significant (r = 0.3732, P > 0.05).

As the cases of SSPE with increased total CSF protein had also increased values of y-globulin, we tried to compare the importance of the contribution of the y-globulin and the other electrophoretic fractions to the total protein concentration in SSPE and in normal CSF. The Sewal-Wright path coeffi- cients underlined the major contribution of y-globulin to total CSF protein in SSPE whereas the other electrophoretic fractions showed a negative influence on CSF proteic bulk (Table 3), which is in agreement with the finding of decreased absolute values of some of these electrophoretic frac- tions. On the other hand, the y-globulin had a minor role in the protein concentration of normal CSF as compared to the other electrophoretic fractions (Table 3).

The y-globulin had an oligoclonal pattern in the CSF in all the cases, whereas in serum 13 (52 %) of the 25 cases studied showed this pattern. Total serum protein and electrophoretic values were not significantly dif- ferent from normal values (Table 1).

The Q-index, the quotient between the relative values of CSF and serum 7-globulins (Schinko & Tshabitscher 1957), was determined in 25 cases. The

17 Acta neurol. scandinav. 63:4

Tab

le I

. Rel

ativ

e va

lues

on

the

elec

trop

hore

sis of C

SF a

nd s

erum

pro

tein

s in

SSP

E pa

tien

ts

CSF

S

erum

Cli

nic

al

To

tal

Pre

- G

lob

uli

n

To

tnl

ras

e

Age

st

ag

e

prot

ein

albu

uiin

A

lbum

in

d-1

c(

2

p ' \

1 pr

otri

n 1

-glo

b.

Q-i

nd-x

(Y-=

Jr.p

) (m

g/lo

O

ml)

%

?6 %

Y

IQ

x z

SS (g

/100

rn

l) :5

1

2 3 4 5 6 7 8 9 10

11

12

13

14

15

16

17

3 10

32

10

12

9 8 9 3 16

11 7 7 11

15

4 16

I1

I1

11s I1

I1

TI

I1

I11 I1

I1

I1

I1

I1

11

If

rr

I1 I

26

17

33

41

53

33

49

23

40

53

23

25

23

34

23

46

23

1.4

7.7

1.9

2 .o 3.3

5.1

2.1

3-7

5.8 4.9

6.4

3.2

3.5

1-7

3-2

4.2 6.5

29.0

32.9

60.9

46.4

34.9

38.1

39.2

78.0

21.3

23.8

46.7

72.2

38.6

47.1

39.0

32.0

36.8

1.8

4.3

6.5

3.6

6.2

14.1

6.

6 6.

0

1.0

11.2

1.

2 4.

8

2.9

5.4

12.0

1.

4

1.7

6.1

13.4

8.

9

4.1

2.3

3.3

5.5

4.5

4.4

7.1

3.1

1.7

4.2

4.3

1.7

3-7

4.3

7.3

5-1

4.1

12.3

9.

0 12

.3

5.4

7.5

9.4

7.3

3.3

7.6

4.0

2.9

3*3

17.9

6.

7 4.

3

1.9

5.5

7.6

2.0

2.0

5.1

7.0

3.9

4.2

4.6

10.7

6.

1

3.4

6.5

9.5

3.7

53.4

+

26.4

+

19.0

+

29.8

+

y.1

+

41.4

+

39.6

+

6.5+

51.4

+

33.6

+

17.1

+

6.9+

25.5

+

34.4

+

40.1

+

38* 3

+

34.0

+

6.6

8.2+

14 .O

t

14.8+

20.1

4.5+

18

.4t

5.0

21.6

+

75.4

+

50.0 - - -

12.6

24.9

+

14.7

t

8.1

5.7

1*4

3 .0 1.6

h)

VI

03

9.2

2.2

1.3

2.4

1*3

0.6 - - - 3.2

1.5

2.3

:

6 i0

17

?3

15

15

15

15 0 16 5 16

?! 37

7 ', ''6

79

31

4s

67

53

41

57

33

31

Ir

. . J

7.5 1.3

3.5

5.3

5.2

1.5 3.9

5.0

7.0

j.2

10.1

1.9

l.5

6 .O

5 .o

4.0

6.3

5.1

5.4

8.1

6.0

6.9

5.7

3.1

7.5

- 0.1

7.7 - 7.1

6.9

3.7

:: .@

7.7

0.6

7.5

0 -6

7 .0

- ".l ? . (1 - 1 .n

1:;

1.7

f .1

4 .0

1.7

3.7

1.7

\D

3.0

h)

v,

Xed

ian

' - O

ligoc

lona

l m

orph

olog

y; N

S -

Not

sig

nific

ant;

(1)

- M

ann-

Whi

tney

U-t

est.

Tab

le 2

. Abs

olut

e va

lues

on

the

elec

trop

hore

sis

of

CSF

and

ser

um p

rote

ins

in S

SPE

pat

ient

s

CSY

(n=

3O)

?-r

um

(1

1-75

)

To

tal

Fre

- G

lob

uli

n

To

tal

prot

chin

al

bu

min

A

lhum

in

41

1%

2 0

< X

pro

tein

g

-qlo

bu

l in

53

1*3

15.1

1

.1

1.7

2.2

1.3

11.1

7.

5 1

.1

6.6

0.4

90

53

1.9

19 07

1

*7

3.5

4.0

2.4

20.4

0

-7

2.1

0.7

0.1

0.5

0.9

1.1

0.7

3.6

Fer

c-n

tilr

10

23

N

Nor

mal

v

alw

s

7.2

0-9

6.5

0.7

97

34

1.5

22.4

1

.3

3.0

4.4

1.4

2.5

7 .e

1.7

1.0

16

.5

0.9

2.c

) 3.0

1.0

1.

7

0.6

12

.1

0.5

1

.3

2.1

0.6

0.9

h’r

dia

n

27

Per

cen

tile

10

19

n

N

Q\

0

CSF

val

ues

are

expr

esse

d in

mg/100

ml

and

seru

m v

alue

s in

g/1

00 m

l. (1

) -

Man

n-W

hitn

ey U

-tes

t; N

S -

Not

sig

nific

ant.

261

median value obtained in SSPE was elevated and compared to the normal median value (2.1 versus 0.5, P < 0.001). The graphic representation of the Q-index shown in Fig. 2 demonstrated that the relative values of y- globulin were greater in CSF than in serum in all except two cases.

Immunoglobulin G, A and M were determined in the last 10 cases studied (Table 4). CSF IgG was over the upper normal value of 4.5 mg/ 100 ml in all cases, with a median value of 14.2 mg/100 ml. Since the median IgG values was normal in serum, the quotient between serum and CSF IgG had to be necessarily decreased. Indeed, we found a value of 99 in SSPE, whereas the normal value is about 400 (Fossan 1977). CSF IgA was doseable in only four of these cases, where the values were slightly above normal. IgM was not detectable in CSF.

DISCUSSION

Our results generally support those previously described in SSPE (Jabbour et al. 1969, Gerson & Haslam 1971, Liano et al. 1971, Link et al. 1973, Vandvik & Norrby 1973, Peter er al. 1974). The selective rise of y-globulin together with a remarkable oligoclonal morphology found in CSF protein

al 10 I 10 20 30 40 50 60

C SF <-Globulin%

Fig. 1. Significative correlation (P < 0.02) between the relative values of y-globulin and the total protein (T.P.) in the CSF from 30 patients with SSPE. Normal values are

delimited by dashed lines.

262

SSPE (nx30) Normal (n-30)

r-glob. (x)

a’ r r = 0.9905

r = 0.9999

r =-0.9887

xz .y

yz .X

w.=

r = 0.0844

r =-0.4716

r = 0.5629

XLa .y

yz .x

xy.2

Table 3. Path coefficients of the y-globulin and of the other fractions (O.F.) on CSF total protein (T.P.) in SSPE and in normals.

electrophoresis, as well as the high values of the Q-index obtained in almost all the cases, are findings suggestive of intrathecal synthesis of y-globulin (Laterre 1965, Cutler et al. 1967, Tourtellotte 1970, Vandvik 1970, Liano ct al. 1971, Link et al. 1973, Vandvik & Norrby 1973, Vandvik & Skrede 1973, Ackermann et al. 1975). The quantitation of immunoglobulins carried out in 10 cases showed this to be almost exclusively constituted of IgG. Unfortunately, the absence of albumin values as determined by immuno- logical methods made it impossible to determine other quotients such as the IgG-index (Delpech & Lichtblau 1972, Christensen et al. 1978).

The characteristic oligoclonal y-globulinic pattern and the absence of transudative features in the CSF protein electrophoresis, the absence of measurable IgM in the CSF of those cases in which immunoglobulins were determined and the normal values of total CSF protein found in the ma- jority of cases, are findings suggesting that there was no significative barrier impairment (Laterre 1965, Tourtellotte 1970, Liano et al. 1971, Link & Miiller 1971, Ackermann et al. 1975, Glasner 1975). Nevertheless, only a detailed study of the CSF proteins could make possible the assessment of the barrier functions in those cases (Clausen 1966, Felgenhauer et al. 1975, 1976). On the basis of the statistical analysis, it is thought that the increased total CSF protein present in four patients (cases 6, 12, 28 and 29) is caused by the great amount of IgG produced in the CNS rather than by a barrier impairment, as suggested to exist in SSPE by Liano et al. (1971) late in the course of the disease. Moreover, as the majority of our cases were in stage 11 at the time of CSF sampling, remarkable protein abnormalities due to barrier breakdown were not to be expected.

An obvious consequence of the great increase of CSF y-globulin was

1

263

6C sp E a n

8 A 5 0

LL In V

0 -

4 0

30

20

10

0

0 0

0

0

0

0

0 . 0 0

0 0

O . 0

0 0 ,'

0 0

0 ,,'

0 ,,' 0

.'

0

10 20 30 40

S E R U M ~ - G I O L W I ~ P a Fig. 2 . Relationship between ihe relative values of CSF and serum y-globulin

(Q-index).

the decrease of the relative values of the other electrophoretic fractions, especially the highest such as albumin, a2- and p-globulins. Moreover, the absolute values of albumin and /%globulin, in spite of the inaccuracy of the method utilized for their calculation, were found to be also decreased in CSF in keeping with previous reports (Liano et al. 1971). This finding sug- gests that the great amount of y-globulin synthesized in the CNS could provoke a real decrease in the concentration of other CSF proteins due, probably, to homeostatic mechanisms for CSF proteins operating by the increase in their reabsorption rate, as described by Davson (1971).

In spite of the severe neuronal and glial damage already present in stage 11, blood-brain and blood-CSF barriers and other mechanisms of homeo- stasis in CSF seem to be maintained in SSPE.

ACKNOWLEDGMENTS We are indebted to Prof. Dr. Joaquim Maia for advice on statistics. We thank Maria de Fa'iima Azevedo and Oscar Campos for technical assistance.

Table 4 . CSF and serum immunoglobulins in I0 SSPE patients

CSP Serum

Case IgG IgA I@ IgC IgA IgL: Serum/CSF IgC

17

22

23

24

25

26

27

28

29

30

16.2

7.5

7.5

15 .o

13.4

22.8

17.7

16.5

5

5 -0

Traces

Traces

0.3

0.3

- 0.3

Traces

- Traces

0.4

640

717

1582

1536

1874

922

819

6 40

947

1638

147

134

218

218

63

161

192

Traces

64

192

325

235

112

142

149

144

174

93

117

114

40

96

184

102

140

40

46

39

186

328

- - Median 14.2 - 935 143 99

Percentiles - - 10 5.1 - 6 40 10 3 40

90 20.3 - 1756 280 257 - -

CSF and serum values are expressed in mg/100 ml.

REFERENCES Ackermann, H. P., H. P. Rieder & R. Wiethrich (1975): Absolute or relative values in

CSF electrophoresis? - An evaluation of the y-globulins in multiple sclerosis and other neurological diseases. Europ. Neurol. 13, 131-143.

Campbell, R. C. (1967): Statistics for biologists, pp. 44-46. Cambridge University Press, Cambridge.

Christensen, O., J. Clausen & T. Fog (1978): Relationships between abnormal IgG in- dex, oligoclonal bands, acute phase reactants and some clinical data in multiple sclerosis. J. Neurol. 218, 237-244.

Clausen, J. (1966): The beta-lipoprotein of serum and cerebrospinal fluid. Acta Neurol. Scand. 42, 153-160.

Cutler, R. W. P., G. V. Watters, J. P. Hammerstad & E. Merler (1967): Origin of cere- brospinal fluid gamma globulin in subacute sclerosing panencephalitis. Arch. Neurol. 17, 620-628.

Cutler, R. W. P., E. Merler & J. P. Hammerstad (1968): Production of antibody by the central nervous system in subacute sclerosing panencephalitis. Neurol. 18 (part 2),

Cutler, R. W. P., G. V. Watters & J. P. Hammerstad (1970): The origin and turnover of cerebrospinal fluid albumin and gamma-globulin in man. J. Neurol. Sci. 10,

129-132.

259-268.

265

Davson, H. (1970): Physiology of cerebrospinal fluid. 2nd ed. Churchill, London. Delpech, B. & E, Lichtblau (1972): Etude quantitative des immunoglobulines G et de

I’albumine du liquide ckphalorachidien. Clin. Chim. Acta 37, 15-23. Ebers, G. C., J. B. Zabriskie & H. G. Kunkel (1979): Oligoclonal immunoglobulins in

subacute sclerosing panencephalitis and multiple sclerosis: A study of idiotypic determinants. Clin. Exp. Immunol. 35, 67-75.

Felgenhauer, K., G. Schliep & N. Rapic (1975): Protein permeability of the blood-CSF barrier. In: Proteins of the biological fluids, Vol. 23, pp. 481-487, ed. Peeters,H.

Felgenhauer, K., G. Schliep & N. Rapic (1976): Evaluation of the blood-CSF barrier by protein gradients and the humoral immune response within the central nervous system. J. Neurol. Sci. 30, 113-128.

Fossan,G.O. (1977): The transfer of IgG from serum to CSF, evaluated by means of a naturally occurring antibody. Europ. Neurol. 15, 231-236.

Gerson, K. L. & R. H. A. Haslam (1971): Subtle immunologic abnormalities in four boys with subacute sclerosing panencephalitis. N. Engl. J. Med. 285, 78-82.

Glasner, H. (1975): Barrier impairment and immune reactions in the cerebrospinal fluid. Europ. Neurol. 13, 304-314.

Jabbour, J. T., J. H. Garcia, H. Lemmi, J. Ragland, D. A. Duenas & J. L. Sever (1969): Subacute sclerosing panencephalitis: A multidisciplinary study of eight cases. J. Am. Med. Assoc. 207, 2248-2254.

Laterre, E. C. (1965): Les protkines du liquide ckphalorachidien B I’ktat normal et pathologique. Arscia, Bruxelles.

Li, C. C. (1955): Population genetics. University of Chicago Press, Chicago, 111. Liano, H., A. Gimeno, M. Kreisler & G. Ramirez (1971): Cerebrospinal fluid proteins in

subacute sclerosing panencephalitis. Acta Neurol. Scand. 47, 579-593. Link, H. & R. Miiller (1971): Immunoglobulins in multiple sclerosis and infections of

the nervous system. Arch. Neurol. 25, 326-344. Link, H., M. Panelius & A. A. Salmi (1973): Immunoglobulins and measles antibodies

in subacute sclerosing panencephalitis (Demonstration of synthesis of oligoclonal IgG with measles antibody activity within the central nervous system). Arch. Neurol. 28, 23-30.

Mancini, G. C. & J. F. Heremans (1965): Immunochemical quantitation of antigens by single radial immunodiffusion. lmmunochemistry 2, 235-241.

ter Meulen,V. & W. W. Hall (1978): Slow virus infections of the nervous system: Virological, immunological and pathogenetic considerations. J. Gen. Virol. 41,

Norrby, E., A. A. Salmi, H. Link, B. Vandvik, J. E. Olsson & M. Panelius (1974): The measles virus antibody response in subacute sclerosing panencephalitis and mul- tiple sclerosis. In: Slow virus diseases, ed. Zeman, W. & E. H. Lennette. Williams & Wilkins, Baltimore.

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

Peter, A,, A. Lowenthal & I. Juvancz (1974): Changes of y-globulins in serum and cerebrospinal fluid of patients with subacute sclerosing panencephalitis. J. Neurol.

Raine, C. S. & B. N. Fields (1974): Neurotropic virus-host relationship alterations due to variation in viral genome as studied by electron microscopy. Am. J. Pathol. 75, 119-138.

1-25.

207, 85-92.

Schinko, H. & H. Tshabitscher (1957): Der y-Quocient als Ausdruck der Relation Liquor zu Serum y-Globulin. Wien. Klin. Wschr. 69, 705-713.

Silva, C. A. & M. J. S i (1978): Electrophoretic pattern of cerebrospinal fluid proteins in non-neoplastic infantile hydrocephalus. Acta Neurol. Scand. 57, 317-328.

Sols, A. (1949): Valoration de las proteinas del suer0 por referencia a la media nor- mal. Rev. Esp. Fisiol. 5, 239-250.

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

Vandvik, B. (1970): Immunological studies in subacute sclerosing panencephalitis. Acta Neurol. Scand. 46 (Suppl. 47).

Vandvik, B. & E. Norrby (1973): Oligoclonal IgG antibody response in the central nervous system to different measles virus antigens in subacute sclerosing panence- phalitis. Proc. nat. Acad. Sci. 70, 1060-1063.

Vandvik, B. & S . Skrede (1973): Electrophoretic examination' of cerebrospinal fluid proteins in multiple sclerosis and other neurological diseases. Eur. Neurol. 9,

Zeman, W. & 0. Kolar (1968): Reflections on the etiology and pathogenesis of sub-

10, 279-304.

224-241.

acute sclerosing panencephalitis. Neurol. 18 (part 2), 1-7.

Received June 26, accepted October 13, 1980

Carlos Alberro Silva, M.D. ServiCo de Neurologia e Neurocirurgia Faculdade de Medicina do Porto Porto Portugal