Herpes simplex virus encephalitis

Journal of the Neurological Sciences, 1984, 63:331-338

331

Elsevier

HERPES SIMPLEX VIRUS ENCEPHALITIS Prolonged Intrathecal IgG Synthesis and Cellular Activity in the Cerebrospinal Fluid with Transient Impairment of Blood-Brain Barrier

EERO TASKINEN l, MARJA-LEENA KOSKINIEMI 2 and ANTTI VAHERI 3

ITransplantation Laboratory, 2Department of Neurology and 3Department of Virology, University of Helsinld, Helsinki (Finland) (Received 1 June, 1983) (Revised, received 1 November,1983) (Accepted 10 November, 1983)

SUMMARY

Alterations of cerebrospinal fluid (CFS) proteins and cells and blood-brain barrier impairment were determined in 4 patients with proven and 2 patients with presumptive herpes simplex virus encephalitis (HSVE) using simultaneous nephelometric measurements of CSF and serum albumin and immunoglobulins and combined Millipore R filtration-cytocentrifuge cytologic techniques. The follow-up period ranged from 17 to 855 days. All patients showed intrathecal IgG synthesis which in 1 case continued for 28.5 months (855 days). The daily production of IgG in the central nervous system ranged up to 1157 mg. CSF-IgA and -IgM were also elevated in the early phase of the disease. The impairment of the blood-brain barrier was variable being apt to develop during the t'n'st 2 months of the disease and diminishing thereafter. Pleocytosis, mainly due to lymphoid cells, varied from slight to severe (325 x 103 cells/ml) and was observed in the CSF of all cases during the first 2 months. Lymphoid reaction (increase of enlarged stimulated lymphoid cells) was persistent and was the most pronounced cellular alteration. The lymphoid reaction and intrathecal IgG synthesis indicated continuous immunoactivation of the CNS, which was most intensive during the first 2 months and appeared to persist for at least 16-28.5 months.

K e y w o r d s : B l o o d - b r a i n barrier - - Cells - - C e r e b r o s p i n a l f l u i d ~ simplex ~

Encephalitis -- Herpes

Proteins

Correspondence: Dr. Eero Taskinen, Transplantation Laboratory, Haartmaninkatu 3, SF-00290 Helsinki 29, Finland. 0022-510X/84/$03.00 © 1984 Elsevier Science Publishers B.V.

332 INTRODUCTION

Intrathecal immunoglobulin synthesis, alterations of the cerebrospinal fluid (CSF) cells and the function of the blood-brain barrier (B-B-B), although reported (Sk01denberg 1972; Dufresne 1973; SkOldenberg et al. 1981) have not been thoroughly studied in the course of herpes simplex virus encephalitis (H SVE). Elevated IgG levels, and elevated albumin concentrations in a few cases, have been reported in the CSF but data on intrathecal immunoglobulin production are rarely available (Roberts-Thomson et al. 1976; Schliep and Felgenhauer 1978; SkOldenberg et al. 1981). Increased cell counts in the CSF have also been detected, but the kinetics of the sequential cellular alterations are not known (Miller et al. 1955; SkOldenberg 1972; Dufresne 1973). These findings are not specific but do provide information on the CNS response in HSVE and assist in diagnosis. In order to define the intensity and duration of the immunoactivation in the CNS of HSVE patients, we have determined the alterations of CSF immunoglobulins and cells as well as B-B-B damage in different phases of 4 proven and 2 presumptive HSVE patients, with prolonged follow-up in 4 cases. PATIENTS AND METHODS

Six patients with HSVE (4 females and 2 males) were treated at the Department of Neurology, University of Helsinki. The age range was 26-68 years. In 4 patients (pts. 1 to 4) the diagnosis was proven by demonstrating HSV or its antigenic material and viral particles (electron-microscopically) in a temporal lobe biopsy. In the other 2 (pts. 5 and 6) the diagnosis was presumed, based on the clinical course and detection of antibodies specific to HSV in CSF and the low serum/CSF antibody titre ratio (< 20). All patients received adenine arabinoside (Vira-A) as an intravenous infusion for 10 days and in patients 2 and 5 this therapy was repeated. In 4 cases (pts 2, 3, 5 and 6) the prolonged follow-up lasted from 507 to 855 days, while in 2 cases the last specimens were obtained 17 and 32 days from the beginning of manifest disease. A clinical account of the patients (excepting pt. 4) has been given elsewhere (Koskiniemi et al. 1980) (Table 1). Ten ml of CSF was withdrawn by lumbar puncture from each patient for cellular and protein determinations. Ten ml of venous blood was taken within 24 h for concomitant serum protein determinations The CSF sample was divided into 2 aliquots. One was used for determination of CSF albumin, IgG, IgM, and IgA levels, together with serum albumin and IgG levels by automated nephelometry (Technicon Autoanalyzer; Taskinen 1983). The CSF albumin level was used as an indicator for the B-B-B function. Local IgG production was estimated by using the CSF-IgG index (Delpech and Lichtblau 1972) and also daily IgG synthesis according to Tourtellotte and Ma (1978). The cytological examination comprised the combined Millipore filtration-cytocentrifugation technique, described in detail elsewhere (Taskinen 1983). Briefly, the total count and the differential count including the relative and absolute numbers (cells/ml) of various cell types were determined.

333 TABLE 1 CSF ALBUMIN (mg/1) AND CELLS (cells/ml) IN HERPES SIMPLEX VIRUS ENCEPHALITIS Patient No./ outcome l/Died (71 d)

2/Died (18 mo)

3/Severe disability (28 mo) 4/Died (30 d)

5/Mild disability (29 mo)

6/Severe disability (21 mo)

Days after onset

Albumin

Total count

Lymphoid cells

MP a

PNL b RBC~

119900 17700

3471 2182

2314 -

+ -

10500

1302

-

+

1052

130

1771

389

4 8 12 32

ND ND 661 157

313 128100 117500 19800

8 15 26 228 248 501

ND ND 181 258 227 139

5350 3438 11 800 313 1299

42 254

1510 460

2160 313

4 5 6 12 17

ND ND ND 222 204

10000 4688 4063 3750 2188

8 13 22 42 55 72 221 483

ND ND ND 3019 ND 616 566 469

275000 325000 61600 24800 34400 11900 26800 4414

15600 25800 8200 24600 3752

9150 8600 3570 2143 662

39 451

200 186

20200 2 886

16000 1934

4248 952

2 850

1920

1370

Upper limit of normal 95~o percentile Mean

1360 + 775

865 + 519

140

+

199

< 1~o

487 + 377

a Mononuclear phagocytes. Polymorphonuclear leukocytes. c Red blood cells.

The total count and the differentiation of CSF

cells i n t o m a i n g r o u p s , i.e.,

l y m p h o i d cells, m o n o n u c l e a r p h a g o c y t e s , p o l y m o r p h o n u c l e a r l e u c o c y t e s , m e n i n g e a l cells, a n d a d d i t i o n a l cells o r a t y p i c a l cells, w e r e e v a l u a t e d f r o m t h e M i l l i p o r e filter s p e c i m e n s . T h e s e p r e p a r a t i o n s w e r e m a d e f r o m e x a c t l y 1.0 m l o f C S F f i x e d i m m e d i a t e l y

334 after collecting in 96 ~o ethanol (1-2 ml) and were stained with the Papanicolau method. Three slides from each specimen were processed in cytocentrifuge by using 0.3 ml of CSF, 800-1000 r.p.m, for 10 rain and M G G staining. The lymphoid cells were divided in these specimens into 3 subgroups: lymphocytes, enlarged lymphoid cells (EL), and plasma cells. EL were defined only on the basis of their light-microscopic structure. No immunological characterization was performed. Most of these cells probably represent stimulated or transformed forms of lymphocytes, with few blast forms. The diameter of EL exceeded 12/~m (see Fig. 2). The nuclear/cytoplasmic ratio was often decreased. The chromatin pattern was net-like, coarse or loose, and one or two nucleoli could usually be seen. The amount of cytoplasm varied and its staining properties ranged from deep basophilia to pale cells. Accordingly, basophilic or pale EL of varying sizes could be seen. A control population was obtained from the out-patient department, University Central Hospital, Helsinki (Taskinen 1983). Forty-six patients in whom no evidence of CNS involvement could be determined served as controls for the cellular examinations of the CSF. The mean age of the control population was 32 years (range 15-65 years). The upper limit of normal for the total count was 2850 and for the EL 400 cells/ml and/or 31 ~o of all the lymphoid cells. If the number of EL exceeded 400 cells/ml an absolute lymphoid reaction was considered to be present and if their proportion was more than 31~o of the CSF lymphoid population. If the number of EL exceeded 400 cells/ml an absolute lymphoid reaction was considered to be present and if their proportion was more than 31 ~o of the CSF lymphoid population, the condition was referred to as a relative lymphoid reaction. The Laboratory of Clinical Immunology, Helsinki, performed the protein determinations and used reference values calculated from a control population.

RESULTS

Protein determinations

The integrity of the B-B-B, assessed using the CSF albumin concentration, varied widely from normal to severely impaired (Table 1). In 3 cases (pts. 2, 4 and 6), no barrier damage was found at all. In patient 1, the initial barrier dysfunction of moderate severity subsided during the first month of the disorder and of the 2 other subjects the initially severe damage was recovered in one and declined in the other. Local production of IgG was observed in at least some phase in all subjects as judged from determinations of daily IgG synthesis (Fig. 1) The IgG index also indicated intrathecal production of this immunoglobulin; in 2 specimens (patient 1/day 12 and patient 3/day 42) the values approached the upper normal limit while daily IgG synthesis was increased. This discrepancy probably is due to the fact that Tourtellotte's formula is not correct for the severe B-B-B damage which is apparent in these patients (1 and 3). During long-term follow-up continuous intrathecal IgG production, with an

335

I000

zoo .'

3

k

3.oo

.z

6o

u.

40

2.00

"-

20 l¢~ J4

1.50 t~

k

1.00 x/e

\"~-~

_

_~

II.

i

i

io

I

i

'

1

I

' ' "lbO

I

I

'

'lOOO

DAYS AFTER ONSET OF SYMPTOMS Fig. 1. Intrathecal IgG synthesis in 6 patients with HSVE. IgG production is demonstrated both as daily synthesis ( e - - e ) and as IgG index (o---o).

increasing trend, was found in 2 patients (pts. 3 and 6) up to 21.5 months (646 days) and 27.5 months (826 days). In an additional subject (pt. 5), production continued to decline but was still abnormal on day 483 as was the IgG index on day 855. The absolute IgG concentration was increased up to 2400 mg/1 in all CSF specimens studied from day 6 in 1 case to day 855 in another except for the last specimen in 1 patient on day 826. Increased concentrations of CSF-IgA and -IgM were found in some specimens from all patients during the first 2 months, as early as day 6. In 1 patient, IgA concentration was still increased 21.5 months after the onset of symptoms.

Cytologicalfindings Pleocytosis was found in most specimens examined during the In'st 2.5 months of disease (Table 1). The increase in cells appeared within 4 days of the onset of symptoms and varied greatly from 3 750 to 325000 cells/ml. Peak values occurred during the first few months followed by a varying decline in the cell counts. In 2 patients elevated cell concentrations were still evident 15-16 months from the onset of disease. In the differential count a shift of varying magnitude towards lymphoid cells was particularly observed in the early specimens. Some neutrophils and mature plasma cells were also detected, but their role in the differential count was not significant. No evidence of destruction of nervous tissue or intracranial haemorrhage, in the form of vacuolated macrophages (lipophages?), erythrophages or siderophages in CSF, was found despite the occurrence of red blood cells in many specimens. An absolute lymphoid reaction, i.e., a pathologically increased number of enlarged stimulated lymphoid cells was observed in all preparations investigated from day 12 to

336

f I

•

Fig. 2. Enlargedlymphoidcells (arrows)among monouclearcells in a specimentaken 248 days (pt. 8) after the onset of the disease. Cytocentrifugepreparations and MGG staining. Magnification x 1200. 15-16 months (Fig. 2). The lymphoid reaction was the most pronounced cellular alteration with levels up to 250 times those of controls. One patient (pt. 5) displayed a second peak of EL on day 221 and a lymphoid reaction was also present on day 248 in patient 2, in spite of a normal total cell count. DISCUSSION The IgG response in CSF was marked as regards both concentration and daily synthesis. In agreement with this, Schliep and Felgenhauer (1978) stated that the most substantial IgG response in the CNS was found in HSVE and neurosyphilis. Furthermore, quantitatively intrathecal IgG production of H SVE greatly exceeded the average daily synthesis of 12 mg in multiple sclerosis or 8 mg in the early phase of aseptic meningitis found in other studies (Tourtellotte and Ma 1978; Taskinen 1983). The differences may be due to varying immunogenic properties of different viruses as well as to various pathogenetic mechanisms. It has been suggested that IgG response in MS is mediated by unspecific polyclonal B cell activation (PBA) (MOiler et al. 1980), while in HSV encephalitis, at least in the early phase, the massive IgG response is thought to represent both PBA and a specific response to virus-induced antigens (Vandvik et al. 1982). The appearance of increased CSF-IgA and -IgM concentrations together with an lgG response is common phenomenon in infectious disorders of the CNS, e.g. in bacterial (Kaldor and Ferris 1969; O'Toole et al. 1971) and in viral meningitides

337 (Roberts-Thomson et al. 1976; Frydrn et al. 1978). Elevated CSF-IgA and -IgM concentrations in our material occurred mainly transiently in the early phase of the disease, i.e. in the same period when B-B-B impairment was also evident. This is in contrast to the persistent, though declining, appearance of HSV-IgM and -IgA antibodies within the CNS during long-term follow-up, as reported by Sk01denberg et al. (1981). These observations, if confirmed, suggest a difference in kinetics of production of specific vs. total CSF-IgM. Even though the CSF protein pattern in HSVE was characterized by the strong and prolonged IgG response, no diagnostically HSVE-specific findings were obtained. Thus the demonstration of infectious virus or viral antigen in brain biopsies and specific serum/CSF antibody determinations are required (Koskiniemi et al. 1980) for the aetiological diagnosis. The cytological alterations were similar to those encountered in aseptic meningitis (Taskinen 1983) as regards cell numbers or lymphoid predominance in the differential count. Such a similarity between disorders with varying aetiologies indicates an unspecific response of the CNS which has been proposed to be mediated in the CSF through reactive meningeal involvement (Hapel 1975; Oehmichen 1976). No characteristic cytological features, e.g., inclusion bodies (Gupta et al. 1972) were found. The persistent lymphoid reaction, even in specimens with a normal cell count, was a phenomenon which has not been previously reported in HSVE. It seemed to follow roughly the fluctuations in CSF total cell count and IgG production and was interpreted to indicate the immunoactivation of the CNS (Taskinen 1983). The increase in enlarged lymphoid cells, up to 250 times control values, was the most prominent alteration in the CSF. On the other hand, the activity of the mononuclear phagocytes in the CSF remained low, in contrast to the situation in the primary brain lesions (Olischer 1971). Similar observations have been made in multiple sclerosis and aseptic meningitis (Olischer 1971; Oehmichen and Treff 1973; Schlote and Roos 1974; Taskinen 1983). The low number of mononuclear phagocytes in the CSF may be due to lymphokinemediated inhibition (Jenssen et al. 1976). Decreased phagocytic ability of these mononuclear cells has also been found in inflammatory disorders (Oehmichen 1978). The treatment with Vira-A did not seem to induce any significant changes in CSF cells or proteins. Thus, determination of CSF cells and proteins do not afford the possibility of estimating the response to antiviral therapy. One patient with fatal outcome had normal barrier function and poor CNS immune response, while the patient with the most favourable outcome showed and enormous intrathecal cellular response and IgG production together with severe barrier impairment. These observations suggest that strong immunoactivation of the CNS may be associated with a favourable outcome in HSVE. A lymphoid reaction with intrathecal IgG production was interpreted as indicating immunoactivation of the CN S (Oehmichen 1976; Link 1979; Taskinen 1983), which probably represents both polyclonal B cell activation and the immune response to H SV-related antigens. These alterations appeared soon after the onset of symptoms and persisted for as long as 16-28.5 months. Both alterations reflect continuous immuno-

338 logical activity o f the C N S . T h i s activity might be the result o f either p e r s i s t e n c e o f viral antigen o r d e v e l o p m e n t o f n e w v i r u s - i n d u c e d e p i t o p e s in the C N S . REFERENCES Delpech, B. and E. Lichtblau (1972) Etude quantitative des immunoglobulines Get de l'albumine du liquide c6phalo-rachidien, Clin. Chim. Acta, 35: 15-23. Dufresne, J.J. (1973) Praktische Zytologie des Liquors, Ciba-Geigy AG, Basle. Fryd6n, A., H. Link and E. Norrby (1978) Cerebrospinal fluid and serum immunoglobulins and antibody titers in mumps meningitis and aseptic meningitis of other etiology, Infect. Immun., 21: 852-861. Gupta, P. K., P.C. Gupta, S. Roy and A.K. Banerji (1972) Herpes simplex encephalitis, cerebrospinal fluid cytology studies - Two case reports, Acta Cytol., 16:563-565. Hapel, A.J. (1975) The protective role of thymus-derived lymphocytes in arbovirus-induced meningoencephalitis, Scand. J. Immunol., 4: 267-278. Jenssen, H. L., H.J. Meyer-Rienecker and H. Werner (1976) Nachweis eines Faktors mit elektrophoretischer Zellmobilit~.tshemmung im Liquor cerebrospinalis bei multipler Sklerose, J. Neurol., 214: 45-59. Kaldor, J. and A.A. Ferris (1969) Immunoglobulin levels in cerebrospinal fluid in viral and bacterial meningitis, Med.J. Aust. , 2:1206-1209. Koskiniemi, M. L., A. Vaheri, V. Manninen, S. Valtonen, L. Ketonen, E. Taskinen, K. Sainio, P. Karli and M. Haltia (1980) Herpes simplex virus encephalitis - New diagnostic and clinical features and results of therapy, Arch. Neurol.(Chic.), 37:763-767. Link, H. (1979) Some aspects of immune reactions of the brain. In: F.C. Rose (Ed.), Clinical Neuroimmunology, Blackwell Scientific Publications, Oxford, pp. 12-28. Miller, J. K., F. Hesser and V.M. Tompkins (1955) Herpes simplex encephalitis - Report of 20 cases, Ann. Int. Med., 64:92-103. M611er, E., H. StrUm and S. AI-Balaghi (1980) Role of polyclonal activation in specific immune responses Relevance for findings of antibody activity in various diseases (Editorial), Scand. J. Immunol., 12:177-182. Oehmichen, M. (1976) Cerebrospinal Fluid Cytology, Georg Thieme, Stuttgart. Oehmichen, M. (1978) Mononuclear phagocytes in the central nervous system. In: Origin, Mode of Distri-

bution, and Function of Progressive Microglia, Perivascular Cells of lntracerebral Vessels, Free Subarachnoid Cells, and Epiplexus Cells, Springer-Verlag, Berlin.

Oehmichen, M. and W.M. Treff (1973) U nterschiedliches postfunktionelles Phagocytose-Verhalten yon Liquorzellen, Verh. Dtsch. Ges. Path., 57: 268-271. Olischer, R. M. (1971) Der Nachweis der unspezifischen Esterase in Liquorzellen - Eine cytodiagnostische Zusatzuntersuchung, Z. NeuroL, 200: 61-69. O'Toole, R.D., G.F. Thornton, M.K. Mukherjee and K.N. Neogy (1971) Cerebrospinal fluid immunoglobulins in bacterial meningitis, Arch. Neurol. (Chic), 25: 218-224. Roberts-Thomson, P.J., M.M. Esiri, A.C. Young and I. Maclennon (1976) Cerebrospinal fluid immunoglobulin quotients - Kappa/lambda ratios, and viral antibody titres in neurological disease, J. Clin. Path., 29:1105-1115. Schliep, G. and K. Felgenhauer (1978) Serum-CSF protein gradients, the blood-CSF barrier and the local immune response, J. Neurol., 218:77-96. Schlote, W. and W. Roos (1974) Gibt es ein charakterisches Liquorzellbild bei multipler Sklerose? Beitrag zur Bedeutung der grossen pyroninophilen Rundzellen in Liquor cerebrospinalis, Nervenarzt, 45: 579-587. Sk61denberg, B. (1972) On the role of viruses in acute infectious diseases of the central nervous system Clinical and laboratory studies on hospitalized patients, Scand. J. Inf. Dis., Suppl. 3: 1-96. Sk61denberg, B., K. Kalimo, A. Carlstrfm, M. Forsgren and P. H alonen (1981) Herpes simplex encephalitis A serological follow-up study of synthesis of herpes simplex virus immunoglobulin M, A and G antibodies and development of oligoclonal immunoglobulin G in the central nervous system, Acta Neural. Scand., 63: 273-285. Taskinen, E. (1983) Cerebrospinal Fluid Cells and Proteins in Neurologic Diseases with Known or Suspected Immunoactivation of the Central Nervous System, Thesis, University of Helsinki. Tourtellotte, W. and B.I. Ma (1978) Multiple sclerosis - The blood-brain barrier and the measurement of de novo central nervous system IgG synthesis, Neurology, 28(2):76-83. Vandvik, B., F. Vartdal and E. Norrby (1982) Herpes simplex virus encephalitis - Intrathecal synthesis of oligoclonal virus-specific IgG, IgA and IgM antibodies, J.Neurol., 228:25-38. -