Oligoclonal gamma-globulin banding of cerebrospinal fluid in patients with subacute sclerosing panencephalitis

Journal of the Neurological Sciences, 1977, 32:395--409

395

© ElsevierScientificPublishingCompany,Amsterdam- Printed in The Netherlands

OLIGOCLONAL GAMMA-GLOBULIN BANDING OF CEREBROSPINAL FLUID IN PATIENTS WITH SUBACUTE SCLEROSING PANENCEPHALITIS Comparison of the Electrophoretic Pattern with that in Multiple Sclerosis and Congenital Infections

H. SIEMES, M. SIEGERT, F. HANEFELD, H. W. KOLMELand F. PAUL Interdisziplin~'re Arbeitsgruppe "Klinische und Experimentelle Plasmaprotein/brschung", Kinderklinik und Neurologische Klinik, Klinikum Charlottenburg, Freie Universitiit Berlin, Neurologische Abteilung der Schlossparkklinik, Berlin (West-Germany)

(Received 29 November, 1976)

SUMMARY Lumbar cerebrospinal fluid (CSF) of 8 patients with subacute sclerosing panencephalitis (SSPE) was examined by agarose gel electrophoresis. In comparison with normal controls and children with different neurological diseases (including infections, tumours and degenerative diseases) the quantitative evaluation of the pherograms by an analog computer revealed an extreme change of the y-globulin profile. All cases showed 6-7 abnormal subfractions consisting of 2-4 tall, markedly protruding spikes and several smaller intermediate fractions. The oligoclonal 7,-globulin contributed 20.142.5 ~ to total protein. This particular ?,-globulin profile seems to be highly indicative of the diagnosis of SSPE. It can be distinguished from the oligoclonal pattern in patients suffering from multiple sclerosis (MS) and congenital infections. The CSF protein profile of 13 patients with MS was different from that in SSPE in that it showed 1-5 monoclonal 7,-fractions in every case with none or only one peak protruding more markedly. The percentage of all subfractions amounted to 4.5-23.8 ~ of total protein. As in MS, the aspect of oligoclonality in 9 children with congenital infections (cytomegalic inclusion body disease, toxoplasmosis and rubella) was quite variable, as again 1-5 abnormal subfractions were detected. Their relative concentrations, on the whole ranging from 0.6-12 ~ of total protein, was considerably lower than in SSPE.

INTRODUCTION Subacute sclerosing panencephalitis (SSPE) is a progressive encephalitis characThis work was supported by the Deutsche Forschungsgemeinschaft.

396 terized by mental deterioration, movement disorders, myoclonic jerks and signs of diffuse neurological dysfunction. The age of onset has been reported to range from 2 to 21 years (Jabbour, Duenas, Sever, Krebs, and Hotta-Barbosa 1972). Death may occur as early as 3 months or as late as 8 years after the onset of the disease; however, most patients die within 1-3 years. In addition to the classical form ofSSPE, a remittent form similar to the course of multiple sclerosis (MS) with exacerbations and remissions has been described (Jabbour et al. 1972). SSPE is a slow viral infection of the CNS, probably caused by the measles virus (Sever and Zeman 1968; Lenette 1975). Though many studies concerning the hostvirus interaction have been performed, the pathogenesis of this aberrant form of infection is unknown (Lenette 1975). Studies of brain tissue and cerebrospinal fluid (CSF) with isotopic (Cutler, Watters, Hammerstad and Merler 1967), immunochemical (Kolfir, Dencker, Obru6n6k, Cernfi and Skatula 1966; Bergman, Dencker, Johansson and Svennerholm 1968, Tourtellotte, Parker, Herndon and Cuadros 1968) and immunohistochemical techniques (Ter Meulen, Enders-Ruckle, Mfiller and Joppich 1969) indicate that immune processes within the CNS are a prominent feature of SSPE. These cause an increase of immunoglobulin G (IgG) concentrations in the CSF. Both paper and cellulose acetate electrophoresis of CSF-proteins therefore show a marked increase of v-globulins in relation to the other proteins as the most consistent pathological finding (Bauer 1961; Van Sande, Karcher and Lowenthal 1961; Llano, Gimeno, Kreisler and Ramirez 1971). In addition to these findings, an oligoclonal banding of the v-globulin has been observed using the technique of agar or agarose gel electrophoresis (Laterre 1964; Lowenthal 1964; Bokonjic et Renders 1964; Laterre, Callawaert, Heremans and Sfaello 1960; Peter, Lowenthal and Juvancz 1974; Bollengier, Delmotte and Lowenthal 1976). The diagnostic significance of this phenomenon remained uncertain for two reasons: (1) oligoclonal v-fractions occur in several other neurological diseases, in particular in MS (Laterre 1964; Lowenthal 1964; Laterre et al. 1970; Link and Miiller 1971; Castaigne, Lhermitte, Schuller, Delasnerie, Deloche and Dumas 1972; Ansari, Wells, and Vatassery 1975); and (2) in most studies published so far, quantitative separation of monoclonal from non-monoclonal Vfractions has not been performed. The object of the present study was to investigate the morphology and quantitative aspects of oligoclonal subfractionation of CSF y-globulins in SSPE by a refined method of agarose gel electrophoresis. In addition the oligoclonal v-globulin pattern of CSF in patients with MS and congenital infections was analysed for comparison with SSPE. PATIENTS AND METHODS Agarose gel electrophoresis of an equal quantity of CSF-protein was performed as described previously (Siemes, Siegert and Rating 1975). For quantitative evaluation of the protein fractions the pherograms were resolved into Gaussian curves by means of a modified analog computer (Du Pont Curve Resolver 310). The Gaussians, generated by the computer, were successively adjusted to the fractions until the superposition of the Gaussians fitted the scanning curve.

397

Fig. 1. Evaluation of electrophoretic mobility of CSF protein fractions. Fixed marks in the baseline on the screen of the analog computer (albumin (A) = 40 and application slot (S) = 100) were adjusted to the actual pherogram. The horizontal position of all protein fractions was then related to this extendable scale (zero located anodically of the prealbumin, the major part of y-globulin cathodically of the application slot).

CSF of 43 normal children and adolescents in whom lumbar puncture was performed to exclude meningitis served as controls. Each regular protein fraction was characterized by its concentration in relation to total protein ( ~ ) and its electrophoretic mobility (Fig. 1). The normal range was defined by the 10th and 90th percentiles. The characterization of the protein fractions by electrophoretic mobility was especially necessary with regard to the )'-trace fraction ()' 4 in Fig. 2) which is a non-IgG protein and otherwise could erroneously be considered a monoclonal IgG band (Bader, Rieder and Kaeser 1973). A total of 1218 patients suffering from different neurological diseases had been studied including CSF samples from 3 children with SSPE and 12 patients with MS who were sent to the pediatric clinics from other hospitals. The evaluation of pathological pherograms was first tried on the basis of the regular ),-globulin pattern and only if necessary for mathematical reasons were new fractions introduced. Monoclonal )'-bands appeared as additional fractions with variable electrophoretic mobility. They were characterized by narrow-based and tall Gaussian-shaped curves in comparison with the regular ),-fraction of identical electrophoretic mobility. They either protruded against the diffuse background of the normal )'z- and )'a-fraction or replaced the regular )'-profile nearly completely. None of the children with acute bacterial meningitis (including tuberculous meningitis) nor the patients with aseptic meningitis nor the children with acute viral encephalitis had a CSF pherogram with monoclonal )'-banding. The same was true for children with hereditary degenerative diseases or tumours of the CNS (exception: 1 child with a medulloblastoma). In all cases of elevated relative concentrations of )'-globulins there had been a polyclonal pattern (Fig. 3). In contrast to these findings, the CSF pherograms of 3 groups of patients suffer-

398

m 610 nm

A

~82 nm

Pc - Nt: 106~ MC. m. IO*/,y s

610 nrn

g82 nro

Pr. - Nr. I 0 6 4 H. C. m. lO*/~ys

F~

P2

A

0.1 ~.i

0.2

0.2' 0.2" 131 132 ~

'Y1 Y2

73

"Y4

Fig. 2. CSF pherogram (A), and analog computer analysis (B), of a child without disease of CNS. Scanning of albumin and globulins at different wavelengths (482 nm and 610 nm respectively); P1 and P2 = prealbumins, A -- albumin, definition of the globulins according to Lowenthal (1964). The proteins of the 71-fraction are unidentified, the 7~- and 73-fractions mainly consist of immunoglobulin G, the protein of the 74-(?-trace)fraction is unrelated to immunoglobulin G. ing from a subacute or chronic form of encephalitis or encephalomyelitis had shown monoclonal I g G populations: (1) All 8 patients with SSPE, consisting o f 4 girls and 4 boys 6-15 years o f age. The diagnosis o f SSPE had been established on the following criteria: the clinical course o f the disease, typical periodic complexes in the electroencephalogram, examination o f the C S F by paper, cellulose acetate or agarose gel electrophoresis, titration o f the measles antibodies in serum and CSF. The clinical data o f the 8 patients are given in Table 1.

399

61Ohm

~82 nm

Pr.-Nr. 1037 B~Th.m. 2~i~y s

610 nrn

1,82 n m

Pr. - Nr. 103 ? B Th.rn. 2~14y s

I°1 1

P2

O.I a,iCtiA a.2 a,2Aa-2'ctLrfll ~2 '~ t i 1 | t

'Y1

'Y2

"Y3 f

'Y4 t

Fig. 3. CSF pherogram (A), and its analysis (B), of a child with acute viral encephalitis showing the polyclonal 7-globulin pattern of acute inflammatory disease of CNS. For explanation of the symbols see Fig. 1; ~', ~, = increase above and decrease below normal range respectively, aiA and a~A = additional plasma protein fractions due to blood-CSF barrier damage. (2) Nine of 27 children with a congenital infection of the CNS, 3 weeks to 4 months of age (5 cases with cytomegalic inclusion body disease, 4 cases with toxoplasmoses and 1 case with rubella). The diagnosis of a congenital infection was established on the basis of the characteristic clinical symptoms (moderate to severe CNS involvement) and on serological findings. In all cases of cytomegalic inclusion body disease the virus was found in the urine. (3) Thirteen patients with MS, 15-34 years of age. The criteria for MS were: clinical evidence of multifocal lesions of the CNS, a typical clinical evolution of the disease and CSF findings (normal or slightly raised cell count and total protein; elevation of IgG in the majority of the cases).

400 TABLE 1 CLINICAL DATA OF 8 CHILDREN WITH SSPE Patients ~ No.

1 2 3 4 5 6 7 8

sex

m f f m f m m f

age (yr)

Duration EEG of findings disease b (periodic complexes)

10 9¼ 14 6 8½ 15½ 11¼ 8

4 mo I yr 8 mo 6 mo 1 yr 2 yr 6yr 1½ yr

Raised measles virus antibody titers ~ serum

i

q I I ~ ,

Total protein (mg/100 ml)

17 0 12 5 4 15 6 3

27.0 28.4 31.6 35.8 26.0 65.0 13.0 22.9

CSF ~

• • t ~

CSF cell count ( x/ram a)

~ 1 ! . : ~

a The patients are grouped in accordance with Table 2. b Before agarose gel electrophoresis was performed. c Haemagglutination-inhibition test.

RESULTS

(1) Subacute sclerosing panencephalitis The cell count and total protein concentration o f the CSF were normal or slightly elevated. Fig. 4. shows the C S F - p h e r o g r a m and corresponding analysis f r o m a child with SSPE. In contrast to the pattern o f increased regular ~,-globulin fractions characteristic of acute purulent or non-purulent inflammatory diseases o f the C N S (Fig. 3), the protein profile o f this child showed 4 narrow-based, highly protruding spikes and 3 smaller intermediate F,-fractions. With the aid o f the analog computer a total of 6-7 abnormal y-subfractions could be separated in all patients, 2-4 markedly protruding and several small intermediate fractions. The normal y-profile (Fig. 2) was completely replaced by irregular fractions cathodically of the ~'1- or ~,~-globulin. The percentage o f the 7-subfractions on the whole ranged f r o m 2 0 . 1 - 4 2 . 6 ~ o f total protein clearly demonstrating that the excessive elevation o f ~'-globulin (21.5-44.4 ~ ) was due to oligoclonal y-globulin production. The results o f the analysis o f all patients are recorded in Table 2. As the regular (~,~-), ys- and ~4-fractions could not be separated by the m e t h o d o f analysis described, it is uncertain whether small intermediate fractions between the highly protruding subfractions (e.g. F'08 in Fig. 4) represent oligoclonal ~,-globulins in every case. They could be caused in part by underlying polyclonal immunoglobulin G. The pathological y-bands showed not only the electrophoretic mobility o f the ~,z- and ~,z-fractions (in normal individuals mainly consisting o f polyclonal i m m u n o globulin G) but also o f F'-trace protein (~'4). M o r e o v e r in 6 out of 8 cases, 1-2 fractions

401 A 610nm

I

[

[1 4 8 2 n m

Pr.-Nn 1424 K.M. f. 14 ys

B

Pr.-Nr. 1,~2,~ K.M. f 14ys

P1

P2

A

(zl ~i

(x2 c¢2"¢¢f'~1 132 •

Y1 YZY0tY02Y0~4Y0s'Y0b'Y01

Fig. 4. CSF pherogram (A), and analog computer analysis (B), of a child with SSPE, demonstrating m a r k e d oligoclonal ~,-globulin banding. F o r explanation of the symbols see Fig. l ; ~', ~, = increase above and decrease below the normal range respectively; ~'01-~'o~ = a b n o r m a l 7-subfractions.

were located even more cathodically. The excessive elevation of y-globulins resulted in a secondary decrease of part of the other protein fractions (Fig. 4). The number of patients was too small to draw conclusions with respect to a correlation between duration of the disease and any particular aspect of oligoclonal banding.

402 TABLE 2 G A M M A - G L O B U L I N PATTERN IN PATIENTS WITH SUBACUTE SCLEROSING PANENCEPHALITIS (GROUPED ACCORDING INCREASING 7-GLOBULIN)

CQses a No.

Elec t r oph oresis b er(figures) ond m r of "y-froctions (o) and pothotoli¢ot ~l " subfroctions (i t/

j:y

~/~

/ Cont rots c

n=23

//~,

1:

"///"/'~

0.9

p.,/

¢ 11&-2.31

'11

V2 ;60

s"

.

5.C//" /5/.,~ 2.2 4,0

//..+./

,./5

~.~

I,&-Z31

I totol cr I~(-subf 1~-lTfOb.

25.2

24.2

25.8

42.5

44.4

--

6.0-10¢

6.9 S/:'f'l,,~'/,/~

+..

12.B-t..4

...2

~.?

0.4 - l_O

V3 ~;o

232

~4 40

do mr

0

Case number according to Table 1. b cr = relative concentration (% of TP), mr = relative mobility. Normal range of regular 7-fractions, cr = figures and mr = hatched fields. a S = application slot of samples.

(2) Multiple sclerosis In 13 patients with MS, the cell c o u n t a n d total p r o t e i n values were n o r m a l or slightly raised (Table 3). T h o u g h the relative c o n c e n t r a t i o n o f ),-globulin was within n o r m a l limits in 3 out o f 13 patients, the ),-pattern was a b n o r m a l in every case. A s in the congenital infections, it was characterized by m o n o c l o n a l b a n d s with a m a r k e d variability in n u m b e r as 1-5 m o n o c l o n a l ),-populations were identified. In cases where several fractions were found, these were d i s t r i b u t e d in the f o r m o f ascending a n d descending steps (Fig. 5A). Typically, the apex o f a single fraction was f o u n d to be only slightly p r o t r u d i n g , with the exception o f one fraction t h a t could a p p e a r m o r e

403 TABLE 3 GAMMA-GLOBULIN PATTERN IN PATIENTS WITH MULTIPLE SCLEROSIS (GROUPED ACCORDING INCREASING 7-GLOBULIN) Patients N

CSF

Electrophoresis d

sex, a g e ,

cr ( f i g u r e s )

*'Jcl=.=w=,~ ccb

,

,.2,,,

,..2,~

,,

I rP c

~/.~///~,

.o

3

~

~7.o

re. ½Yl-

28

48.0

I

39.5

13

&,.O

4 yr

re.

5

m, 28yr lyr

re,

° , . +

lyr

re,

7

8

f, 24yr" r e lyr

,, ~3~

5"

t a t a r cr

/

/

~/A"/,"o: S'~'a;/// ;~

~i'2-

,

/

z: ~//+//~/

~7]~/

y-subt.lv-t+tob.

,~

~,/.~

~//////~/

~.9 •

33.0

53.0

f, 15+yr re, 3yr

28

39.5

f, 25~r r e , 2~Yrt"

23

48.0

11

m, 3t.yr re-chr, 9yr

1

2Z5

12

fo 2~yr r e , ~yr

16

20.5

13

f, 29yr re, ½yr

12

68.5

, +~/~.~

1.0

0.7

////(

,..,

,:

,~ 2.2~ ,

,

45

7.7

4.0

8.I

+,3

9.9

1.5

1 2.4

9.8

14.4

13.3

15.6

13.4

17.0

16.2

20.7

18.1

21.2

17.3

21.6

19.9

21.8

22.6

25.4

23.8

27.4

( (.(/

/'~6

20

10

" y - f r a c t i o n s (o)

,,

1;yr.

9

re,

0+5

~+/, ~ /

ch~. ;½~r

a n d m r of

,,,d po,,oto++~,,t "y-suhf,.,ctJo,,sr.)

,:.3

///

//,

.

1.4-2,3

y/~

y..'/// 6,0-10.0

/////

Controls e

n=23 u

(~

sr

mr

(~

a re = remittent attacks, chr = chronic progressive, re-chr = primary remittent-chronic progressive; duration o f disease. b Cc = cell count ( x / m m a ) . e Tp = total protein (mg/100 ml). d cr = relative concentration (% o f TP), mr = relative mobility. e N o r m a l range o f regular ;,-fractions, Cr = figures and mr = hatched fields. S = application slot o f samples.

404

~82 nm

610 nrn

A

Pr. - Nr, 159J K.M. f 15ys

P1

1

P2

A

0-1 OLi

1

0-2

0.2' OL2" 131 132 1;

1

'Y1 "Yo1'YO2"YO3'YO4~'OS"Y4

1

t

B

!'11 Pr.- Nr I~ 72 M. O. m. 2f/zrns

/ 0.2 (x2A (x2' o~2" Pt ~2

A

1

i

i

t

#

"Y1

Y2 Ym Ye2 Ye3"YO~¥4 t

Fig. 5. Analysis of CSF pherograms of a child with MS (.4), and congenital toxoplasmosis (B). For explanation of the symbols see Figs. 2-4.

405 markedly. The relative concentration of all of these monoclonal fractions taken together varied from 1.5-23.8 ~ of total protein; they thus predominantly formed the y-globulin of the CSF in most of the patients (Table 3). In 11 out of 13 cases, the electrophoretic mobility of these fractions was restricted to the yz-y4-region of the pherograms; in 2 patients they were located cathodically of y-trace. There seemed to be no apparent correlation between the duration of the disease and any particular aspect of oligoclonality.

(3) Congenital infections of the CNS In the majority of the 9 children with different congenital infections, the CSF TABLE

4

GAMMA-GLOBULIN PATTERN IN INFANTS WITH CONGENITAL (GROUPED ACCORDING INCREASING 7-GLOBULIN)

Patients | ielo ~ o

No. la~ln,=t=a 1

CSF C¢ b

I rD °

f,&wk ¢yt.

11

2

f, 3 w k

6

57.0

3

f, 4 m o cyt.

1

25.4

4

f, 5wk ¢yt.

26

~0.3

t, 4wk

13

153.0

6

m, 2 mo IOX,

33

26.0

7

f, 4wk

63

38.2

f, 4wk

46

/*5.5

f, 2mo

16

5

8

9

oyt.

cyt,

tOX.

tox.

rub.

INFECTIONS

Ele c troph oresis d Cr(flgures) and mr of ~/-fro¢#oni

(e) and pothMe~col ~- subfraeff~ns(#

-sub~ [y- glob,

128.0

V,&..,'>~"

112.0

~× f//× Controls e

( n : 2 0 , age: 3 w e e k s 4 months)

B/

Vl I~0 s~

0,6-1,6

0.6

11.8

0.7

12.0

,5.0

12.4

1.3

13.2

0.6

15.3

10.2

13.7

12.5

14.5

7.4

16.5

0.6

t 6.9

--

2.8-7.9

/////> 1.0 -

3.1

0,3 .

1.2

/_///

Y/_///~

/////.@

v2

~'3

v,;

,;o

Iio

6o mr

" ~)

C y t . = cytomegalic inclusion body disease, t o x . = t o x o p l a s m o s i s , r u b . = r u b e l l a . b C c = cell count ( × / m m a ) . e T p = total protein ( r a g / 1 0 0 ml). d cr = relative concentration ( % o f T P ) , m r = relative mobility. e Normal range of regular y-fractions, cr = figures and m r = hatched fields. S = application slot of s a m p l e s .

406 cell count and total protein values were normal, and in some were only moderately elevated (Table 4). The relative concentration of y-globulin was raised above normal in all cases. A variable number of 1-5 abnormal y-subfractions was found. In the child with rubella only 1 fraction and in 5 children with cytomegalic inclusion body disease 1 or 2 of these y-fractions appeared, whereas in all cases of toxoplasmosis 4-5 y-subfractions could be detected (Fig. 5). In contrast to SSPE, the relative concentration of monoclonal y-fractions was considerably lower, ranging from 0.6-12.5 % of total protein. If 1 or 2 monoclonal y-fractions appeared, these were distributed in the region between the regular y2- and y4-fraction. In the case of several monoclonal bands these were located in the ys- and y4-position or even cathodally in relation to the y-trace. Oligoclonal y-bands could be demonstrated in infants only until 4 months of age (some older patients investigated before the age of 2 years showed elevated y-globulin fractions characteristic of a polyclonal immunological reaction). DISCUSSION Both paper and cellulose acetate electrophoresis are useful routine methods tor determining abnormal relative y-globulin concentrations in CSF. However, agar and agarose gel electrophoresis proved superior to these techniques in revealing multiple monoclonal y-bands in certain neurological diseases, especially in subacute sclerosing panencephalitis (SSPE) and multiple sclerosis (MS). The demonstration of the oligoclonal banding caused by local IgG synthesis within the CNS is considered by some authors to be a more significant finding than the mere elevation of y-globulins (Laterre et al. 1970; Castaigne et al. 1972; Ansari et al. 1975). There are a few other neurological diseases where monoclonal y-subfractions have been observed, although less frequently: these include congenital infections in children (Siemes, Siegert and Lison 1974), neurosyphilis (Laterre et al. 1970) and occasionally in tumours and degenerative diseases of the CNS in adults (Laterre et al. 1970; Peter et al. 1974). The appearance of monoclonal bands amidst a diffuse background or even replacing the whole normal y-globulin pattern is thought to be due to a restricted heterogeneity of the immune response (Heremans and Masson 1973). According to these authors this phenomenon could be due to immunological immaturity (as in congenital infections) or prolonged stimulation caused by persistence of antigens (as in SSPE). The significance for the pathogenesis of these diseases is unknown, but in infants with congenital infections of the CNS the immunological pattern seems to vary depending on the nature of pathogenic organisms. The production of immunoglobulin G within the CNS of patients with SSPE is probably due to an intense stimulation of mononuclear cells which have been attracted to the CNS by virus antigens (Vandvik 1973). Results of serological investigations in oligoclonal IgG bands isolated by preparative electrophoresis from CNS or CSF specimens suggest that these fractions represent antibodies directed against different viral antigens (Link, Panelius and Salmi 1973; Vandvik 1973). In our study, with the aid of analog computer analysis, several distinct oligoclonal y-fractions could be demonstrated in all children, pointing to an equal number of antigens.

407 MS brain tissue may contain oligoclonal IgG (Link 1972). As mononuclear cells present in the CSF of MS patients are able to synthesize monoclonal IgG (Sandberg-Wollheim 1974) it is uncertain whether CSF IgG derives from damaged brain tissue (Olsson, Link and Miiller 1976) and if so to what extent. The demonstration of monoclonal IgG in all patients in our study points again to the significant utility of the oligoclonal pattern in establishing the diagnosis of MS. In the present investigation children with congenital infections showed oligoclonal ),-globulins up to 4 months of age. It is unknown how long this CSF-pattern persists. Recent reports have demonstrated that viral agents other than measles virus can cause the clinical picture of SSPE. In those cases, school age children with chronic progressive panencephalitis of late onset were found to be due to congenital rubella infection simulating SSPE (Townsend, Baringer, Wolinsky, Malamud, Mednick, Panitsch, Scott, Oshiro and Cremer 1975; Weil, Itabashi, Cremer, Oshiro, Lennette and Carnay 1975.) Although CSF ),-globulin was markedly increased in these patients, the authors did not observe oligoclonal banding, which can probably be explained by the methods used for their electrophoretic studies. The demonstration of oligoclonal ),-globulin banding in the CSF by quantitative agarose gel electrophoresis is a highly valuable diagnostic test for SSPE for three reasons: (1) In all patients with SSPE, a striking subfractionation of )'-globulin was observed. Only the 8 children with SSPE (out of 1203 patients with different neurological diseases) showed 2-4 markedly protruding monoclonal ),-globulin peaks. Together with small intermediate fractions, there were as many as 6-7 abnormal )'-subfractions, their summed percentage ranging from about 20-43 ~ of the total protein. (2) The strong immunological response revealed by CSF electrophoresis is not always in accordance with low haemaglutination-inhibition titres against measles virus antigens which have been found in some CSF samples and brain extracts (Bernard, Ripert, Haddad, Riberi and Depieds 1974). This discrepancy could be the consequence of stimulation either by other virus antigens of or binding of the CSF antibodies which are identified by the haemaglutination-inhibition test to antigens (Riberi, Bernard and Depieds 1975). Therefore the demonstration of excessive amounts of oligoclonal )'-bands could be as valuable for establishing the diagnosis as is the determination of raised measles virus antibody titres in patients with SSPE. (3) In comparison with SSPE, the oligoclonal pattern of CSF )'-globulins in MS and in congenital infections was different. In cases of MS there were 1-5 monoclonal fractions distributed in stepwise fashion with none or only one band considerably protruding. With regard to the quantitative aspect of),-patterns there was a slight overlap of MS with SSPE. In MS, the percentage of the )'-subfractions all together ranged up to 23 ~ of total protein compared with 20-43 ~o in SSPE. In congenital infections the ),-profile differed considerably with respect to the number and aspect of fractions (1-5 subfractions, only one peak markedly protruding) as well as the percentage of total protein (0.6-12.5 ~o). From the present investigations the following conclusions can be drawn: monoclonal ),-globulin bands can be demonstrated in all patients with SSPE and MS

408 a n d in a p r o p o r t i o n of children with congenital infections by a refined m e t h o d of analyzing agarose gel electrophoresis. F u r t h e r m o r e the shape of the y-profile a n d q u a n t i t a t i o n of the m o n o c l o n a l bands also allow certain distinctions between these three types of disease. ACKNOWLEDGEMENTS The authors particularly t h a n k Prof. B. Stiick, K i n d e r k l i n i k , Rudolf-VirchowK r a n k e n h a u s , Berlin, a n d Prof. F. Dressier, K i n d e r k r a n k e n h a u s , Berlin-Neuk6lln, for access to records a n d they gratefully acknowledge the expert technical assistance of Mrs. R. F r o m m h o l d a n d Mr. R. Jaroffke.

REFERENCES Ansari, K. A., K. Wells and G. T. Vatassery (1975) Quantitative estimation of cerebrospinal fluid globulins in multiple sclerosis, Neurology (Minneap.), 25: 688-692. Bader, R., H. P. Rieder and H. E. Kaeser (1973) Die Bedeutung der diskontinuierlichenZonierung des Immunglobulinbereichs f~ir die Diagnose neurologischer Erkrankungen, Z. NeuroL, 206: 25-38. Bauer, H. (1961) Cerebrospinal fluid findings in panencephalitis. In: L. van Bogaert, J. Radermecker, J. Hozay and A. Lowenthal (Eds.). Encephalitides, Elsevier, Amsterdam, pp. 675-680. Bergman, L., S. J. Dencker, B. G. Johansson and L. Svennerholm (1968) Cerebrospinal fluid ~'-globulin in subacute sclerosing leucoencephalitits, J. Neurochem., 15: 781-785. Bernard, D., G. Ripert, A. Haddad, M. Riberi and R. Depieds (1974) An immunological study of subacute sclerosing panencephalitis - - Measles and tissue antigens and immunoglobulins,Ann. ImmunoL (Inst. Pasteur), 125 C: 461-469. Bokonjic, R. and J. Renders (1964) 15tude 61ectrophor6tique de 400 liquide c~phalorachidiens pathologiques, Acta neurol, belg., 64: 769-776. Bollengier, F., P. Delmotte and A. Lowenthal (1976) Biochemical findings in multiple sclerosis, Part 3 (Immunoglobulinsof restricted heterogeneity and light chain distribution in cerebrospinal fluid of patients with multiple sclerosis), J. NeuroL, 212: 151-158. Castaigne, P., F. Lhermitte, E. Schuller, N. Delasnerie, G. Deloche and Y. Dumas (1972) Oligoclonal aspect of gamma-globulins in CSF - - Diagnostic value. In: E. J. Field, T. M. Bell and P. R. Carnegie (Eds.) Multiple Sclerosis (Progress in Research, Vol. 3), North-Holland Publishing Company, Amsterdam, London, pp. 152-156. Cutler, R. W. P., G. V. Watters, J. P. Hammerstad and E. Merler (1967) Origin of cerebrospinal fluid gamma-globulin in subacute sclerosing leucoencephatitis, Arch. Neurol. (Chic.), 17: 620-628. Heremans, J. F. and P. L. Masson (1973) Specific analysis of immunoglobulins, Clin. Chem., 19: 294-300. Jabbour, J. T., D. A. Duenas, J. L. Sever, H. M. Krebs and L. Horta-Barbosa (1972) Epidemiology of subacute sclerosing panencephalitis (SSPE) - - A report of the SSPE Registry, J. Amer. med. Ass., 220: 959-962. Kol~r, O., S. J. Dencker, M. Obru6n6k, J. Cern~ and Z. Skatula (1966) Zur Bedeutung der ~-Globulinfraktion im Hirngewebe bei der subakuten sklerotisierenden Leukoencephalitis, Dtsch. Z. Nervenheilk., 188 : 222-223. Laterre, E. C. (1964) Les Protdines du Liquide Cdphalo-rachidien tJ l'Etat Normal et Pathologique, Arscia, Brussels. Laterre, E. C., A. Callawaert, J. F. Heremans, and Z. Sfaello (1970) Electrophoretic morphology of gamma-globulins in cerebrospinal fluid of multiple sclerosis and other diseases of the nervous system, Neurology (Minneap.), 20: 982-990. Lenette, E. H. (1975) Summary of the Conference on Cellular Immunity and SSPE, Arch. NeuroL (Chic.), 32: 489-493. Liano, H., A. Gimeno, M. Kreisler and G. Ramirez (1971) Cerebrospinal fluid proteins in subacute sclerosing panencephalitis, Acta neuroL scand., 47: 579-593.

409 Link, H. (1972) Oligoclonal immunoglobulin G in multiple sclerosis brains, J. neurol. Sci., 16: 103-114. Link, H. and R. MOiler (1971) Immunoglobulins in multiple sclerosis and infections of the nervous system, Arch. Neurol. (Chic.), 25: 326-344. Link, H., M. Panelius and A. A. Salmi (1973) Immunoglobulins and measles antibodies in subacute sclerosing panencephalitis, Arch. NeuroL (Chic.), 28: 23-30. Lowenthal, A. (1964) ,4gar Gel Electrophoresis in Neurology, Elsevier, Amsterdam, New York. Olsson, J.-E., H. Link and R. M011er (1976) Immunoglobulin abnormalities in multiple sclerosis, J. neuroL Sci., 27: 233-245. Peter, A., A. Lowenthal and I. Juvancz (1974) Changes of ~'-globulins in serum and cerebrospinal fluid of patients with subacute sclerosing panencephalitis, J. Neurol., 207: 85-92. Riberi, M., D. Bernard and R. Depieds (1975) Evidence for the presence of 2 chain dimers in cerebrospinal fluid of patients suffering from subacute sclerosing panencephalitis, Chin. exp. lmmunol., 19: 45-53. Sandberg-Wollheim, M. (1974) Immunoglobulin synthesis in vitro by cerebrospinal fluid cells in patients with multiple sclerosis, J. lmmunol., 3: 717-730. Sever, J. L. and W. Zeman Eds. (1968) Conference on measles virus and subacute sclerosing panencephalitis, Neurology (Minneap.), 18: 1-192. Siemes, H., M. Siegert and H. Lison (1974) Liquorelektrophorese bei connataler Encephalitis, Mschr. Kinderheilk., 122: 440--442. Siemes, H., M. Siegert and D. Rating (1975) Das Liquorproteinprofil normaler Kinder und seine Abh~ingigkeit vom Lebensalter, Neuropiidiatrie, 6: 383-397. TerMeulen, V., G. Enders-Ruckle, D. M011er and G. Joppich (1969) Subacute sclerosing panencephalitis - - Virological and immunohistological studies, Acta neuropath. (BerL), 12: 244-259. Tourtellotte, W. W., J. A. Parker, R. M. Herndon and C. N. Cuadros (1968) Subacute sclerosing panencephalitis - - Brain immunoglobulin G, measles antibody and albumin, Neurology (Minneap.), 18: 117-126. Townsend, J. J., J. R. Baringer, J. S. Wolinsky, N. Malamud, J. P. Mednick, H. S. Panitsch, R. A. T. Scott, L. S. Oshiro and N. E. Cremer (1975) Progressive rubella encephalitis, New Engl. J. Med., 292: 994--998. Vandvik, B. (1973) Immunopathological aspects in the pathogenesis of subacute sclerosing panencephalitis with special reference to the significance of the immune response in the central nervous system, Ann. elin. Res., 5: 308-315. Van Sande, M., D. Karcher and A. Lowenthal (1961) Electrophoretic observations in subacute sclerosing leucoencephalitis. In: L. van Bogaert, J. Radermecker, J. Hozay and A. Lowenthal (Eds.), Encephalitides, Elsevier, Amsterdam, pp. 684-689. Weil, M. L., H. H. Itabashi, N. E. Cremer, L. S. Oshiro, E. H. Lenette, and L. Carnay (1975) Chronic progressive panencephalitis due to rubella virus simulating subacute sclerosing panencephalitis, New. Engl. J. Med., 292: 994-998.