Journal of the Neurological Sciences, 1982, 54:209-226
Elsevier Biomedical Press
HERPES SIMPLEX ENCEPHALITIS An Immunohistological Study of the Distribution of Viral Antigen within the Brain
MARGARET M. ESIRI
Department of Neuropathology, Radcliffe Infirmary, Oxford OX2 6HE (Great Britain) (Received 24 September, 1981) (Accepted 29 October, 1981)
An immunoperoxidase technique was used to map the sites of herpes simplex virus antigen (VA) within the brain in 29 autopsied cases of herpes simplex encephalitis. Attention was directed particularly to those parts of the brain that are known from pathological studies to be involved in the disease. Material was studied from cases surviving for varying periods from a few days to a few years after the onset of neurological disease. VA was found within the brain in all cases dying within 3 weeks of onset, but in none dying thereafter. VA was already most abundant in patients dying within the first week and remained plentiful during the first 16 days. Inflammation and necrosis reached a peak when detectable virus was waning. VA was concentrated mainly in the medial and inferior temporal lobes, hippocampus, amygdaloid nuclei, olfactory cortex, insula and cingulate gyrus. It was invariably present on both sides of the brain but was more abundant on one side than the other. Virus was found in glial cells of the olfactory tracts but not in relation to trigeminal pathways. Attention is drawn to focally extensive infection of the granule cells of the dentate fascia, and the possible significance of this finding is discussed.
Herpes simplex virus (type 1) (HSV1) is recognised to be the commonest identifiable cause of sporadic fatal encephalitis in humans (Meyer et al. 1960; Rappel et al. 1971; Baringer 1978). The disease is characterised by the involvement predominantly of those localised regions of the brain comprising the limbic system This work was supported by a grant from the Medical Research Council. 0022-510X/82/0000-0000/$02.75 © Elsevier Biomedical Press
210 TABLE 1 SUMMARY OF PATIENTS Case No.
Most severely damaged cerebral hemisphere
Acute onset of headache followed by confusion, stiff neck and fever. CSF lymphocytosis. HSV cultured from right temporal lobe biopsy taken shortly before death.
Acute onset of fever, confusion and dysphasia followed by fits and coma. CSF pteocytosis (60~ lymphocytes). HSV cultured from left frontal brain biopsy.
Acute onset of confusion and left sided weakness followed by fits and coma. CSF contained 4 lymphocytes only. HSV grown from right frontal lobe biopsy.
Few days pyrexia and malaise followed by fits, confusion, drowsiness, mild left hemiparesis, and later coma. Lymphocytosis of CSF. HSV cultured from brain at autopsy.
Sudden onset of fits and aphasia with pyrexia. Mild CSF lymphocytosis.
Sudden onset of fits followed by coma.
Acute onset of fits followed by confusion and later coma.
Acute onset of confusion and pyrexia followed by fits and coma. HSV cultured from brain biopsy.
"Flu'-like illness for 1 week followed by photophobia and coma. Lymphocytosis of CSF. HSV cultured from brain biopsy and at autopsy.
Malaise for 2 weeks followed by sudden fever and confusion, later coma. Lymphocytosis of CSF.
1 t days
Acute onset of pyrexia and confusion followed by dysphasia and right hemiparesis. Lymphocytosis of CSF. HSV cultured from brain at autopsy.
Acute onset of headache, neck stiffness and photophobia followed by fits and coma. CSF lymphocytosis.
"Flu"-like illness for a few days followed by confusion, drowsiness and dysphasia. Later fits and coma.
Acute onset of headache and confusion with pyrexia. Later coma. Lymphocytosis of CSF.
Acute confusion, headache and drowsiness followed by coma.
Acute onset of confusion and pyrexia followed by fits and coma. Mild lymphocytosis of CSF. 2 week's malaise and headache followed by drowsiness and pyrexia ; later fits and coma. CSF pleocytosis, chiefly neutrophils. Acute onset of headache and pyrexia with confusion and dysphasia. Later drowsiness and coma. CSF lymphocytosis.
Headache and vomiting for several days followed by fits and left hemiplegia. Later comatose.
211 Table 1 (continued) Case No.
Most severely damaged cerebral hemisphere
24 days 6 wks
Multiple sclerosis for many years (confirmed at autopsy). Sudden onset ofdysphasia and hemiparesis, followed by coma. "Flu"-like illness for few days followed by confusion, slight neck stiffness and later coma. Lymphocytosis of CSF. Malaise with headache and vomiting followed by confusion, slight neck stiffness and later coma. Acute onset of confusion followed by coma. 3 days' malaise followed by pyrexia, fits, left hemiparesis and coma. CSF lymphocytosis. Never regained consciousness. CSF HSV titre 1/32. Acute encephalitic illness with survival for 3 months from onset, in comatose state. Head injury when aged 20 followed by post traumatic epilepsy. Sudden onset of malaise and odd behaviour followed by aphasia and pyrexia. CSF pleocytosis, 60~ lymphocytes. HSV cultured from left temporal lobe biopsy. Later profoundly demented. Acute onset of confusion, headache and pyrexia. Normal CSF. Later coma for 3 weeks, followed by slow improvement. Rise in serum HSV titre from 1/32 to 1/512 between 1st and 3rd weeks of illness. Later demented and epileptic. Pyrexial illness with confusion followed by coma lasting several weeks. CSF lymphocytosis. Later demented and dysphasia, incontinent and with a right hemiparesis. Unsteady with slurred speech for a few days followed by pyrexia, neck stiffness and coma for 1 month. CSF pleocytosis. Later severe amnesia and blindness.
Right Right Left Right
( H a y m a k e r et al. 1958; A d a m s a n d M i l l e r 1973). These include the o l f a c t o r y p a t h w a y s , a m y g d a l o i d nuclei, septal nuclei, m e d i a l t e m p o r a l lobes, i n c l u d i n g the h i p p o c a m p i a n d p a r a h i p p o c a m p a l gyri, a n d cingulate gyri. A n e x p l a n a t i o n for this l o c a l i s a t i o n o f the p a t h o l o g y o f HSV1 encephalitis has been s o u g h t in t e r m s o f the r o u t e b y which the virus m a y r e a c h the brain. I n p a r t i c u l a r , suggestions o f intran e u r a l s p r e a d o f virus either a l o n g the o l f a c t o r y nerves a n d tracts ( J o h n s o n a n d M i m s 1968; H u g h e s 1969; T w o m e y et al. 1979; D i n n 1979, 1980) o r a l o n g b r a n c h e s o f the t r i g e m i n a l nerve i n n e r v a t i n g the b a s a l meninges ( D a v i s a n d J o h n s o n 1979) have been p u t f o r w a r d . A c c e s s o f virus via the o l f a c t o r y r o u t e w o u l d better explain the l o c a l i s a t i o n o f the lesions since the l i m b i c system includes the central o l f a c t o r y c o n n e c t i o n s , b u t a t t e m p t s to d e m o n s t r a t e virus b y i m m u n o f l u o r e s c e n c e o r culture in the o l f a c t o r y b u l b s o r tracts have o n l y r a r e l y met with success ( M a c C u l l u m 1973; D a v i s a n d J o h n s o n 1979; D i n n 1980). T h e earlier view that HSV1 encephalitis is a
212 manifestation of primary infection by the virus seems untenable, as the incidence of herpetic cold sores prior to the onset of the disease is similar to that in the general population (Leider et al. 1965) and those patients who have a history of recurrent cold sores must be presumed to harbour a lantent infection. The present study is an attempt to shed light on this question by the use of the immunoperoxidase technique to demonstrate sites of viral antigen (VA) in autopsied cases of herpes simplex encephalitis. Although this technique has already been used to demonstrate the presence of virus in samples of temporal and frontal lobes in this disease (Benjamin and Ray 1975; Kumanishi and Hirano 1978), there has been no attempt at systematic mapping of the sites of VA at different stages of the disease. MATERIALS A N D METHODS
Post mortem material, collected over a 20-year period, routinely fixed in formalin and embedded in paraffin wax, from the brains of 29 patients was studied. The diagnosis had been confirmed by viral culture from material obtained at brain biopsy or at autopsy in 13 cases and in the remaining cases the diagnosis was based on pathological findings at autopsy with, in some cases, confirmatory serological evidence of a rising titre of antibody to HSV in serum and/or cerebrospinal fluid. When the diagnosis had been made during life most patients had been treated with either idoxuridine or cytosine arabinoside. Brief details of the patients and the length of their survival from the onset of neurological disease are given in Table 1. Cases 28 and 29, with long survival, were the subjects of a previously published report (Hierons et al. 1978, cases 1 and 2, respectively), and the pathology of cases 1, 2, 3 and 26 has been previously described by Hughes (1969; cases 1, 2, 3 and 5, respectively). When possible, the following regions from each side of the brain were examined: olfactory bulbs and tracts; sites of entry of lateral olfactory striae; septal area and anterior commissure; amygdaloid nuclei; hippocampus, subiculum, entorhinal area and inferior temporal gyrus; insula; cingulate gyrus; post central gyrus (as an example of cortex relatively far removed from limbic connections); pons at the level of entry of the trigeminal nerves; and medulla. In several cases the orbital gyri, subcallosal gyri and trigeminal ganglia were also examined. Serial 6-/~m sections from blocks of these areas were exposed to a rabbit antiserum to HSV1 (anti-HSV) (1/40 dilution) or to control serum (normal rabbit serum or anti-HSV serum pre-absorbed with HSVl-infected fibroblasts). The rabbit antibody to HSV1 was prepared as described in Tomlinson et al. (1974). Sites of VA were detected using the peroxidase anti-peroxidase (PAP) technique, details of which have been published elsewhere (Esiri et al. 1976), substituting the rabbit anti-HSV1 for the anti-immunoglobulin antiserum described in that paper. Control autopsy material subjected to treatment with the same anti-HSV and control sera included sections from the hippocampus and temporal lobe from the brains of 30 patients with other disorders including subarachnoid haemorrhage,
213 cerebral infarction, cerebral tumour, head injury, and Alzheimer's disease. Five surgical temporal lobectomy specimens were also examined in the same way. A semi-quantitative estimate of the amount of VA was recorded on a scale 1-3. A score of 1 was recorded if virus was present in single or in occasional cells in approximately 40 fields using x 25 objective; of 2 if several cells in the same area contained VA and a score of 3 if numerous cells within this area contained VA (with a score of 3 there were often hundreds of cells containing VA). RESULTS
Fifteen of the patients were male and 14 were female. Their ages ranged from 11 to 81 years. The survival times of the patients ranged from 5 days to 3 years from the onset of neurological disease, the onset being identified by excessive drowsiness, headache and/or focal neurological signs and/or epilepsy. This development had in some cases been preceded by a variable prodromal period with less specific complaints (Table 1). Half the patients died within 15 days, and all but 6 within 24 days. The remainder survived for 6 weeks, 3 months, 15 months, 17 months, 2 years and 3 years, respectively (Table 1). All the cases show pathological change affecting both sides of the brain but in all cases it was clear that one cerebral hemisphere was more severely affected than the other. In 17 patients the most severely affected side was the left and in 12 patients it was the right. VA was demonstrated in the brains of all patients dying within 19 days of the onset of neurological disease but in none dying thereafter. The prevalence of VA in the different regions of the brain is recorded in Table 2. In the following account the findings are described with respect to patients dying at different times after the onset of illness. First week
Four patients died within the first week of onset of their neurological disease. In their brains there was already an abundance of VA present within both cerebral hemispheres, but slightly more antigen was detected on the more severely affected side. Thus, the mean percent of the maximum possible score for VA achieved on the worse affected side was 51~ and on the less affected side was 33~ (Table 3). On the worse affected side not only was there more antigen but it was also more widely distributed, there being VA at 86~o of the sites examined compared with 65~ on the less affected side. The highest concentrations of virus present were found in the amygdaloid nuclei, cortex and white matter at the sites of entry of the lateral olfactory striae, entorhinal cortex, subiculum, hippocampus (pyramidal neurons and dentate fascia), insula and cingulate gyrus. Small amounts of VA were found in 3 of the 4 cases in which these were available for examination, in scattered glial cells in both olfactory tracts, and in one of two olfactory bulbs. In the dentate fascia of the hippocampus VA was abundant in 1 case and scanty in 3 cases in the
214 TABLE 2 VIRAL ANTIGEN
SCORES FOR PATIENTS
N o viral antigen was discovered in cases 2 2 2 9 in the total of 166 sites examined.
Sites examined for VA
Case No. Week I
1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 2 0 2 1
R o l f a c t o r y bulb
L olfactory bulb R olfactory tract L olfactory tract Cortex at entry R lat. olf. stria C o r t e x at entry L lat. olf. stria R Septal nuclei L septal nuclei R ant. commissure L ant. commissure R amygdala L amygdala R dentate granules L dentate granules R hippocampal pyramids L hippocampal pyramids R e n t o r h i n a l cortex L e n t o r h i n a l cortex R alveus L alveus R subiculum L subiculum R inferior temporal cortex L inferior temporal cortex R cingulate gyrus L cingulate gyrus R insula L insula R post central cortex L post central cortex pons
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
.... 1 . . . . . . 0 . . . . . . . . 0 1 1 1 1 0 0 1 1 0 1 1 1 1 001 1 2 3 3 0 0 1 3 3 - 0 3 - 2 1 1 1 1 0 0 000 1 0 0 0 0 100 1 - 1 0 0 0 01 0 - 0 0 2 113 1 1 3 3 2 0-2 2 2 3 2 3 1 3 1 0 2 2 3 0 3 0 21 0 1 0 0 2 1 0 3 12 1 1 3 31 0 1 0 11 0 12 0 31 1 0 11 1 2 3 3 2 2 3 0 2 3 0 3 3 1 1 3 3 I 21 101 2 0 2 0 1 1 11 0 3 0 2 0 2 3 1 13 3 2 0 3 0 2 1 0 3 3 0 1 3 31 21 0 1 2 3 2 3 020 2 1 2 1 0 2 123 1 2 2 3 0 1 3 3 2 2 1 3 0 3 0 3 1 3 2 3 I 2 3 3 0 - - 0 1 3 2 3 2 2 0 312 0 0 2 0 0 320 1 0 2 0 0 000 0 0 0 0 0 011 0 0 0 0 0 - - 0 0 0
1 30 0 3 31 3 0 30 2 0 3 3 3 1 3 3 3 3 3 3 3 0 2
. . -
0 0 0 2 3
1 3 1 3 3 1 2 2
0 - 0 0 0 O 0 0 0 2 0 0 3 0 0 3 0 0 0 2 0 0 - 0 0 0 0 0 1 0 00 0 2 - 0 2 0 2 01t2 3 2 0 0 3 0 1 1 0 0 l 1 3 1 1 0 0 0 2 3 0 0 3 1 3 3 1 0 1 0 3 3 0 0 0 0 1 0 0 1 0 2 0 0 0 3 0 3 3 1 1 1 0 3 3 0 0 3 I 2 3 0 0 1 0 3 3 0 1 0 0 0 3 2 3 0 0 0 3 2 3 0 0 3 1 3 0 0 2 2 3 0 0 10 0 0 1 3 0 0 0 0 0 0 0 1 0 0(1 0 0 0 0 0
0 1 - 2
. 0 0 3 3 2 3 0 2 2 3 1 3 0 3 2 3 I 3 2 3 2
0 0 -
granule cells o f the more severely damaged hemisphere, but was not seen in the contralateral dentate fascia. In grey matter VA was present in neurons and glial cells (astrocytes and oligodendrocytes) in nuclei and/or cytoplasm. In white matter VA was relatively scanty in glial cells of the olfactory tracts, anterior commissure and alvei o f the hippocampi, but much more abundant in white matter of temporal lobes, insula and cingulate gyri. In none of these cases was VA found in the pons or medulla but in 2 cases VA was present in post central cortex. Pathologically, at this stage o f the disease, inflammation was relatively scanty and most c o m m o n l y seen as perivenous cuffs around veins in the superficial cortex
2 1 2 2 1 3 2 1 1 2 3 3 2 2 3 3 0
215 TABLE 3 C O M P A R I S O N O F SCORES A N D SITES O F VA, E X P R E S S E D AS P E R C E N T A G E S OF POSSIBLE T O T A L F O R CASES W I T H M O S T C O M P L E T E D A T A
25 33 49 24
24 52 65 64
56 79 73 50
80 86 100 79
5 8 9 10 11 14
33 13 39 39 73 42
36 40 36 56 47 81
60 20 77 77 93 75
60 73 77 92 67 100
16 17 18 19 20 21
64 14 19 30 2 51
69 12 19 21 6 52
85 43 31 45 6 85
100 29 31 55 19 86
1st Week 1 2 3 4 Mean 2ndWeek
and in meninges, particularly within sulci adjacent to areas of cortex containing VA. The commonest changes within the brain were shrinkage and eosinophilia of neuronal cell cytoplasm with appearances suggestive of acute ischaemia, together with marked congestion, capillary dilatation and petechial haemorrhages in affected cortical regions. Intranuclear and intracytoplasmic inclusion bodies were occasionally seen and appeared to contain VA, but were inconstant. Many of the shrunken "ischaemic" neurons contained VA.
Second week Ten patients died between days 8 and 14 of their neurological illness. In 6 of these cases most of the regions of interest were available for examination but in the other 4 material was confined to the temporal lobes. Again VA was found in all the brains and in both cerebral hemispheres. When the 6 more extensively mapped cases were compared with the cases dying in the first week there was slightly less antigen overall in the brain and it was slightly less widely distributed. (Table 3). As in the first week, slightly more VA was found in the more severely affected hemisphere, though the differences in the amounts of VA between the two hemispheres was not as great as in the first week. In 2 cases more VA was found in the less than in the more severely affected hemisphere. In these cases the more severely affected hemisphere showed particularly extensive necrosis and inflammation. In 4 of the 6 cases in which they were examined, VA was found in the olfactory tracts, and in 3 of these 4 it was present on both sides (olfactory bulbs were not available for study from any of these cases). VA in olfactory tracts was relatively scanty, and present in glial cells (Fig. 1). The inferomedial temporal lobe, hippocampus and amygdala were sites of abundant VA, or were extensively damaged in these cases. Other sites with plentiful VA at this stage were the posterior orbital and subcallosal cortex, cingulate cortex, cortex adjacent to the entry of the lateral olfactory striae, and insula. In 3 of these 6 cases in which it was examined the post central gyrus contained VA, and in 3 of 6 cases there was also scanty VA in glial cells in the base of the pons within the pontine nuclei. The anterior commissure
Fig, 1. Case 5. Right olfactory tract, longitudinal section, treated with anti-HSV serum. Scattered glial nuclei show positive staining for antigen (arrows). Counterstained with haematoxylin, × 250.
217 contained abundant VA in glial cells in 2 cases and small amounts of VA in 2 of the other 3 cases in which it was examined. The septal nuclei did not contain VA but in one case the septal region was destroyed on both sides and in another case was destroyed on the worse affected side. There was no obvious difference in the extent of damage as between the medial and lateral septal nuclei. In 2 cases there were small foci of VA-containing ependymal cells lining the inferior horn of the ventricle. The distribution of.VA within the temporal lobe and hippocampus requires more detailed consideration. Information was available for at least some of these regions from both sides of the brain from all 10 cases. The inferior temporal gyrus, entorhinal cortex, subiculum and hippocampus contained VA on one or both sides in all cases. In most cases VA was very abundant in cortex in neurons and glia, and in glial cells in underlying white matter (Figs. 2 and 3). In the hippocampus the pyramidal neurons contained VA on one or both sides in all but 1 case and in that case these neurons had been destroyed (case 1 1). Perhaps the most remarkable finding related to the granule cells of the dentate fascia (Figs. 4-8). These cells contained VA in 9 of 10 cases; they contained VA only on the worst affected side in 4 cases, and on both sides in 5 cases. In 3 cases the hippocampi were examined in coronal sections at several levels and it was clear in these cases that more VA was found at the anterior pole of the hippocampus than at its posterior pole. In 7 of the 9 cases with VA in dentate granule cells there was involvement of cells within the entire ribbon present in a coronal section. In the other 2 cases the infection appeared to be restricted to 1 segment; for example the medial extremity in case 11 (Fig. 6). Electron microscopy was performed on blocks of tissue taken
Fig. 2. Case 16. Section of cingulate cortex treated with anti-HSV serum. Heavy infection of neurons and glial cells evident. Counterstained with haematoxylin, × 180.
Fig. 3. Case 6. Section of inferior temporal lobe deep cortex and white matter treated with anti-HSV serum, Heavy infection of glial cells with HSV evident. Counterstained with haematoxylin, x 580.
Fig. 4. Case 14. Coronal section of left hippocampus treated with anti-HSV serum. Uniformally heavy infection of granule cells of dentate fascia. Counterstained with haematoxylin, × 20.
Fig. 5. Case 14. Serial section to Fig. 4, stained with anti-HSV serum preabsorbed with herpes simplexinfected cells. Staining for HSV seen in Fig. 4 is abolished. Counterstained with haematoxylin, × 20.
Fig. 6. Case 11. Coronal section of left hippoeampus treated with anti-HSV serum. Medially situated granule cells of dentate fascia are infected with HSV. Counterstained with haematoxylin, × 30.
Fig. 7. Case 15. Coronal section of right hippocampus treated with anti-HSV serum. Heavy, but slightly patch? infection of granule cells of dentate fascia. Some neighbouring glial cells and a few neurons to left are also infected. Counterstained with haematoxylin, x 26.
Fig. 8. Case 15. Higher power view of focus of infection within the layer of granule cells of the dentate fascia showing staining of dendrites as well as cell bodies. Counterstained with haematoxylin, × 300.
221 from the dentate fascia of case 14 in which the whole ribbon of granule cells contained VA (Fig. 4) and confirmed the presence of intranuclear and intracytoplasmic particles with morphological appearance of HSV. VA was frequently seen in glial cells in the alvei, but only rarely in the fornices and flmbriae. Histologically at this stage there was well advanced oedema and necrosis of many of the areas examined, with loss of cortical neurons, extensive accumulation of lipid phagocytes and well marked perivascular cuffing with lymphocytes and plasma cells, particularly in the cortex and overlying meninges. Most VA was found at sites where these histological changes were least well developed, and often the most necrotic areas contained no VA or only very occasional cells faintly stained for VA. The olfactory tracts showed little or no inflammation even in those cases containing VA in glial cells, although underlying meninges were often inflamed. Third week
Seven patients died between days 15 and 19 of their neurological illness. In these patients' brains VA was found in all cases and was as prevalent in the three cases dying on day 15 as in the cases dying in the second week. Later in this third week less VA was found (Table 2). As expected, the same areas of the brain were involved as in the earlier cases, and inflammation and necrosis were even more advanced. Taking the findings in all cases dying in the third week together (but excluding case 15 for which information is limited to the temporal lobes), VA was found on the worse affected side of the brain at 53~o of the sites examined and the score for VA was 30~ of the total possible. Corresponding percentages for the less affected hemisphere were 49~o and 30~ (Table 3). The total amount of VA was therefore less than in the previous 2 weeks and was also more evenly distributed between the two cerebral hemispheres. Again the areas of brain containing most VA were the temporal lobes, hippocampus, insulae and cingulate gyri. The amygdala, subcallosal gyrus and olfactory cortex rarely contained VA but were often necrotic. Within the temporal lobes the inferior temporal gyrus, entorhinal cortex and subiculum frequently contained VA in neurons and glia, and within the hippocampus there was again severe involvement of the granule cells of the dentate fascia, with VA present in 4 cases and extensive destruction of this layer of cells in the other 3 cases (Table 2). The olfactory tracts were examined in 6 cases and VA was found in glial cells within them in 2 of these cases. Two cases contained VA in post central cortex and 1 case contained VA in glial cells within the base of the pons. Generally the necrosis and inflammation reached a peak intensity in the third week. In some cases the olfactory tracts contained foci of perivascular inflammation whether or not VA could be demonstrated within them. One brain from an elderly patient (case 16) exceptionally showed very little inflammation anywhere though VA was abundant, particularly within the temporal lobes. (This patient also showed
222 no rise in antibody titre in serum or cerebrospinal fluid during the course of her illness). Fourth week onwards One patient died on the 22nd day of the illness, another on the 24th day and another at 6 weeks. The 5 remaining patients died months or years later. No VA was found in any of these cases. Severe necrosis and inflammation of limbic structures, together with prominent reactive astrocytosis, were present in the cases dying within a few weeks of the onset. Later on severe tissue destruction, gliosis, and some residual mononuclear inflammation remained. No VA or other abnormalities were seen in the 3 pairs of trigeminal ganglia examined from cases 16, 18 and 20. No VA was found in the control material except for 1 case of Alzheimer's disease in which a tiny focus of VA was found in the granule cell layer of the dentate fascia of the hippocampus. The significance of this finding is uncertain; staining of an adjacent section with pre-absorbed antibody gave negative results but this could have been due to the absence of the corresponding cells from this section. DISCUSSION
In this study I have attempted to follow the distribution and abundance of VA within the brain during the evolution of herpes simplex encephalitis. The greatest abundance of VA, already widely distributed, was found at the earliest stage at which the patients succumbed, within the first week of onset (It is recognised that timing of the length of the neurolgical illness may be imprecise, since the onset of the disease may be gradual and the neurological illness preceded by a variable prodromal period.) Also during the first week there was found to be the greatest difference in amounts of VA between the two cerebral hemispheres. Statistical analysis failed to show significant differences, but the numbers of patients compared was too small for satisfactory analysis. The fact that the difference in amounts of VA between the two hemispheres was greater in the first than in subsequent weeks may suggest that the disease originates in one hemisphere and then spreads to the other. Even within the first week there was spread of VA beyond the structures comprising the limbic system, as evidenced by the finding of VA in post central cortex in a few cases. However, much the greatest concentrations of VA were found within limbic structures at this stage as well as later on in the disease. During the 2nd and 3rd weeks VA remained abundant in the limbic system on both sides of the brain, slightly more so in the hemisphere showing greatest pathological changes. At the end of the third week VA abruptly disappeared from both hemispheres. The pathological changes were most flamboyant when the amounts of VA were already waning. The impression gained is of a rapidly spreading wave of viral infection within limbic structures, probably starting on one side of the brain and spreading within it and to the other side, lasting about 3 weeks and leaving in its wake a
223 trail of devastatingly severe necrosis and inflammation of infected parts of the brain. VA was found in all types of glial cells as well as in neurons. Oligodendrocytes and astrocytes were frequently involved, but ependymal cells were seen to contain VA in only 2 cases, and in small focal areas only. In experimental infection of mice, in contrast, astrocytes were described as being the only glial cell type readily infected (Townsend and Baring~r 1978). It is difficult to extrapolate from this study to the likelihood of a brain biopsy revealing the diagnosis, as most of the sites examined are relatively inaccessible to the surgeon; but it seems clear that the earlier a biopsy is performed the greater will be the chance of obtaining identifiable virus. The present study indicates that areas of most severe destructive and inflammatory change are not necessarily the areas most likely to contain virus, particularly after the first week of the neurological disease. In clinical studies in which brain biopsy was undertaken there appears to have been a very high incidence of positive viral identification (Whitley et al. 1977, 1981), though the wisdom of undertaking this procedure is still questioned (Caplan 1977; Barza and Pauker 1980; Braun 1980). As virus is widely distributed within the limbic system even at the earliest stage of the disease for which autopsy is available, it is difficult to be sure about where it might have made its first appearance. However the following points are worth making in this connection: (1) The evidence from this study suggests that the main focus of infection initially lies in one anterior temporal lobe, hippocampus, amygdaloid nucleus complex, cortex close to entry of the lateral olfactory stria, insula and cingulate gyrus. There is no clearly discernible progression from one to another of these structures, but virus does also reach the same regions on the other side of the brain at an early stage. This study therefore appears to lend support to the view that the source of the infection is the olfactory pathway. VA was indeed found in the olfactory tracts in 9 of the 15 ocases in which it was sought during the first 3 weeks of the disease; in both olfactory tracts in 8 cases and in one olfactory tract in only 1 case. It should be noted that VA was also simultaneously present within the brain, and this evidence cannot be considered as uniquely favouring spread of virus along olfactory pathways into the brain; it could equally well indicate spread of virus from the infected brain. Inflammation within olfactory tracts was usually scanty. In the older patients there were frequent degenerative changes including the presence of many corpora amylacea, but these are a common finding even in the absence of neurological disease and do not furnish evidence about the pathogenesis of herpes simplex encephalitis. Relatively greater success in demonstrating HSV in olfactory tracts in this study than in earlier ones is likely to be due to the greater sensitivity of the immunoperoxidase (PAP) technique than that of immunofluorescence in demonstrating small amounts of antigen (Sternberger 1979). (2) No evidence was found to support the view that virus may reach the brain via the trigeminal pathway (Davis and Johnson 1979). No VA was found in the meninges at the base of the brain, trigeminal ganglia or root entry zones of the trigeminal nerves. Scanty VA was occasionally found in glial cells in the region
224 of the pontine nuclei in the 2nd and 3rd weeks, and inflammatory cuffing was common in the base of the pons. It seems likely that VA reaches these sites in the pons by spread along neural pathways provided by temporo-pontine and fronto-pontine fibres. (3) Heavy infection of the granule cells of the dentate fascia of the hippocampus was one of the most striking findings in this study. In some cases the whole length of this ribbon of cells, cut in the coronal plane, was involved, while in other cases one segment only was affected. The infected cells were indistinguishable from their uninfected neighbours in routinely stained sections in which all these cells commonly showed severe shrinkage and eosinophilia. There was more extensive infection of these cells at the anterior pole of the hippocampus than further back in the few cases in which this structure was examined at different levels. One possible explanation for this distribution of VA is that virus, once it has reached this layer of cells, spreads very rapidly and extensively within it. It is easy to imagine that this might be the case, as rapid cell-to-cell spread would seem very likely in a layer of close-packed cell bodies such as these, These cells have a rich afferent input from the entorhinal cortex which was also found to be heavily infected, giving ready access via neural pathways to the dentate fascia granule cells. Recent studies (Kosel 1980) have indicated connections in rats between the olfactory bulb and the entorhinal cortex, so that anatomical connections may well permit spread of virus from olfactory tracts to granule cells of the dentate fascia. It would therefore seem possible that herpes simplex virus may lie latent within the olfactory bulbs. (4) Another possibility worth considering is that the dentate granule cells themselves may provide a site of viral latency from which reactivated virus could spread to cause encephalitis. The unexpected finding of 3 cells in this layer apparently containing VA in a patient with Alzheimer's disease could conceivably be a manifestation of latent infection and warrants further investigation (Esiri 1982). The afferent connections of the dentate granule cells with the entorhinal cortex and mossy fibre connections with hippocampal pyramidal cells, which in turn project to the subiculum and contralateral hippocampus, would seem to provide an appropriate anatomical substratum to account for the invariably observed bilateral but asymmetrical distribution of the virus. The relatively slight involvement of the septal nuclei, particularly the lateral nuclei, is difficult to explain on this basis, but may be related to the only slight infection, of long, compact white tracts, which may hamper virus spread. The early clinical features, which may include personality, behavioural and memory disturbances would not be inconsistent with this suggestion. Establishment of HSV1 latency in sensory ganglia is now supported by evidence in both humans (Bastian et al. 1972; Baringer and Swoveland 1973; Plummer 1973; Rodda et al. 1973; Warren et al. 1978) and experimental animals (Stevens and Coon 1971; Walz et al. 1974; Price et al. 1975; Puga et al. 1978; Kristensson et al. 1979; Openshaw et al. 1979; McLennan and Darby 1980). Latency of HSV in brain has been demonstrated occasionally in experimental infections (Knotts et al. 1973; Cook and Stevens 1976; Cabrera et al. 1980), and preliminary
evidence for the presence of the HSV genome in the brain has also been provided in humans (Sequiera et al. 1979) but requires further substantiation. The present demonstration of HSV in striking amounts in the granule cells of the dentate fascia in herpes simplex encephalitis directs attention to these cells and their connections as a possible site of latency from which infectious virus might occasionally reappear, and commends this layer of cells for more detailed studies both in man and animals. ACKNOWLEDGEMENTS
I am most grateful to Dr. A.H. Tomlinson who prepared the HSV1 antiserum, and to Drs. B. Brownell, I. Janota, D.R. Oppenheimer and J.T. Hughes for material from cases of herpes simplex encephalitis. Expert technical help was provided by Mrs. P. Hobday. REFERENCES Adams, H. and D. Miller (1973) Herpes simplex encephalitis - - A clinical and pathological analysis of 22 cases, Postgrad. reed. J., 49: 393-397. Baringer, J. R. (1978) Herpes simplex infections of the nervous system. In: P. Vinken and G.W. Bruyn (Eds.), Handbook of Clinical Neurology, Iiol. 34 (Infections of the Nervous System, Part II), Elsevier, Amsterdam, pp. 145-159. Baringer, J.R. and P. Swoveland (1973) Recovery of herpes-simplex virus from human trigeminal ganglions, N. Engl. J. Med., 288: 648-650. Barza, M. and S. G. Pauker (1980) The decision to biopsy, treat or wait in suspected herpes encephalitis, Ann. intern. Med., 92: 641-649. Bastian, F.O., A.S. Rabson, C.L. Yee and T.S. Tralka (1972) Herpes virus hominis - - Isolation from human trigeminal ganglion, Science, 178: 306--307. Braun, P. (1980) The clinical management of suspected herpes virus encephalitis - - A decision-analytic view, Amer. J. Med., 69: 895-902. Benjamin, D. R. and C. G. Ray (1975) Use of immunoperoxidase on brain tissue for the rapid diagnosis of herpes encephalitis, Amer. J. clin. Path., 64: 472-476. Cabrera, C.V., C. Wohlenberg, H. Openshaw, M. Rey-Mendez, A. Puga and A.L. Notkins (1980) Herpes simplex virus - - DNA sequences in the CNS of latently infected mice, Nature (Lond.), 288 : 288-290. Caplan, L.R. (1977) Ara A for herpes encephalitis (Letter), N. Engl. J. Med., 297: 1288-1289. Cook, M. L. and J. G. Stevens (1976) Latent herpetic infections following experimental viraemia, J. gen. Virol., 31 : 75-80. Davis, L. E. and R. T. Johnson (1979) An explanation for the localisation of herpes simplex encephalitis? Ann. Neurol., 5: 2-5. Dinn, J.J. (1979) Distribution of herpes simplex virus in acute necrotising encephalitis, J. Path., 129: 135-138. Dinn, J.J. (1980) Transolfactory spread of virus in herpes simplex encephalitis, Brit. med. J., II: 1392. Esiri, M.M., C.R. Taylor and D.Y. Mason (1976) Application of an immunoperoxidase method to a study of the central nervous system - - Preliminary findings in a study of human formalin-fixed material, Neuropath. appl. Neurobiol., 2: 233-246. Esiri, M.M. (1982) In preparation. Haymaker, W., M. G. Smith, L. van Bogaert and C. de Chenar (1958) Pathology of the viral disease in man characterised by nuclear inclusions. In: W. S. Field and R. L. Blattner (Ed.), Viral Encephalitis, C. C. Thomas, Springfield, IL, pp. 95-104. Hierons, R., I. Janota and J.A.N. Corsellis (1978) The late effects of neerotising encephalitis of the temporal lobes and limbic area - - A clinico-pathological study of 10 cases, Psychol. Med., 8: 21--42. Hughes, J.T. (1969) Pathology of herpes simplex encephalitis. In: C.W.M. Whitty, J.T. Hughes and
226 F.O. MacCallum (Eds.), Virus Diseases and the Nervous System, Blackwell Scientific Publications, Oxford, pp. 29-37. Johnson, R.T. and C.A. Mires (1968) Pathogenesis of viral infections of the nervous system, N. Engl. J. Med., 278: 23-30, 84-92. Knotts, E.P., M.L. Cook and J.G. Stevens (1973) Latent herpes simplex virus in the central nervous system of rabbits and mice, J. exp. Med., 138 : 740-744. Kosel, K.C. (1980) Olfactory bulb projections to parahippocampal areas in the rat. In: 93rd Annual Session of the Amer. Association o f Anatomists, Nebraska, p. 102A (Abstr.). Kristensson, K., B. Svennerholm, L. Persson, A. Vahlne and E. Lycke (1979) Latent herpes simplex virus trigeminal ganglionic infection in mice and demyelination in the central nervous system, J. neurol. Sei., 43: 253-264. Kumanishi, R. and A. Hirano (1978) An immunoperoxidase study of herpes simplex virus encephalitis, J. Neuropath. exp. Neurol., 37: 790-795. Leider, W., R. L. Magoffin, E. H. Lennette and L. N. R. Leonards (1965) Herpes simplex encephalitis Its possible association with reactivated latent infection, N. Engl. J. Med., 273: 341-347. MacCullum, F.O. (1973) Discussion, Postgrad. med. J., 49: 406-409. McLennan, J.L. and G. Darby (1980) Herpes simplex latency - - The cellular location of virus in dorsal root ganglia and the fate of the infected cell following virus activation, J. gen. Virol., 51 : 233 243. Meyer, Jr. H.M., R.T. Johnson, I.P. Crawford, H.E. Dascomb and N.G. Rogers (1960) Central nervous system syndromes of "viral" aetiology - - A study of 713 cases, Amer. J. Med., 29: 334-347. Openshaw, H., L. V. Asher, C. Wohlenberg, T. Sekizawa and A.L. Notkins (1979) Acute and latent infection of sensory ganglia with herpes simplex virus - - Immune control and virus reactivation, J. gen. Virol., 44: 205-215. Plummer, G. (1973) Isolation of herpes viruses from trigeminal ganglia of man, monkeys and cats, J. infect. Dis., 128: 343-348. Price, R.W., B.J. Katz and A. L. Notkins (1975) Latent infection of the peripheral ANS with herpes simplex virus, Nature (Lond.), 257: 686-688. Puga, A., J.D. Rosenthal, H. Openshaw and A.L. Notkins (1978) Herpes simplex virus DNA and mRNA sequences in acutely and chronically infected trigeminal ganglia of mice, Virology, 89: 102-111. Rappel, M., M. Dubois-Dalcq, S. Sprecher, L. Thiry, A. Lowenthal, S. Pelc and J.P. Thys (1971) Diagnosis and treatment of herpes encephalitis - - A multidisciplinary approach, J. neurol. Sci., 12 : 443-458. Rodda, S., I. Jack and D.O. White (1973) Herpes simplex virus from trigeminal ganglion, Lancet, i: 1395-1396. Sequiera, L.W., L.C. Jennings, L.H. Carrasco, M.A. Lord, A. Curry and R.N.P. Sutton (1979) Detection of herpes simplex viral genome in brain tissue, Lancet, ii: 609-612. Sternberger, L.A. (1979) Immunocytoehemistry, 2nd edition, John Wiley and Sons, New York, NY. Stevens, J.G. and M.L. Cook (1971) Latent herpes simplex virus in spinal ganglia of mice, Science, 173: 843-845. Tomlinson, A.H., I.J. Chinn and F.O. MacCallum (1974) Immnno-fluorescence staining for the diagnosis of herpes encephalitis, J. elin. Path., 27 : 495~,99. Townsend, J.J. and J.R. Baringer (1978) Central nervous system susceptibility to herpes simplex infection, J. Neuropath. exp. Neurol., 37: 255-262. Twomey, J. A., C. M. Barker, G. Robinson and D. A. Howell (1979) Olfactory mucosa in herpes simplex encephalitis, J. Neurol. Neurosurg. Psychiat., 42: 983-987. Walz, M.A., R.W. Price and A.L. Notkins (1974) Latent ganglionic infection with herpes simplex virus types I and lI - - Viral reactivation in vivo after neurectomy, Science, 184:1185-1187. Warren, K.G., S.M. Brown, Z. Wroblewska, D. Gilden, H. Koprowski and J. Subak-Sharpe (1978) Isolation of latent herpes simplex virus from the superior cervical and vagus ganglions of human beings, N. Engl. J. Med., 298: 1068-1069. Whitley, R.J., S.-J. Soong, R. Dolin et al. (1977) Adenine arabinoside therapy of biopsy-proven herpes simplex encephalitis - - National Institute of Allergy and Infectious Diseases Collaborative Antiviral Study, N. Engl. J. Med., 297: 289-294. Whitley, R.J., S.-J. Soong, M.S. Hirsch, A.W. Karchmer, R. Dolin, G. Galasso, J.K. Dunnick and C.A. Alford (1981) Herpes simplex encephalitis - - Vidarabine therapy and diagnostic problems, N. Engl. J. Med., 304: 313-318.