Disseminated Herpesvirus (Herpes simplex) · 2 C.M. 8 1-I Mouth and Classical florid disseminated 60-r 22/1954 months nose herpesvirus infection Fatty liver 3 B.M. 5-27 Notrecorded

J. clin. Path., 1972, 25, 79-87

Disseminated Herpesvirus hominis (Herpes simplex)infection: retrospective diagnosis by light andelectron microscopy of paraffin wax-embeddedtissuesA. G. ROSE AND W. B. BECKER'

From the Departments of Pathology and of Bacteriology and MRC Virus Research Unit, University of CapeTown Medical School

SYNOPSIS Disseminated Herpesvirus hominis infection had previously been reported from otherparts of the world, but the disease was initially recognized at necropsy in Cape Town in 1957(McKenzie, Hansen, and Becker, 1959).The present retrospective study established that the disease had been occurring before 1957 but

that its aetiology had not been recognized. An awareness of the disease, together with an adequateknowledge of the clinical history, the morbid anatomy, and histology of the lesions, usually allows a

confident diagnosis to be made.

Hass (1935) described a case of disseminated Herpessimplex infection with necrotic lesions of the liverand adrenal in a premature infant. Since then therehave been numerous reports of disseminated diseasein neonates and also in older children.The association of disseminated herpesvirus

infection with malnutrition in children after theneonatal period was first recognized in Cape Townin 1957 (McKenzie et al, 1959; McKenzie, 1961).If this disseminated disease caused by the ubiquitousHerpesvirus hominis (Becker, 1966) occurred for thefirst time in 1957 it would suggest precipitatingfactors such as the emergence of a more virulentstrain of the virus. An alternative explanation wouldbe that similar cases were seen before 1957, but thattheir viral aetiology was not appreciated at the time.The postmortem findings and the histology in theflorid cases are distinctive enough (McKenzie et al,1959; Becker, Naude, Kipps, and McKenzie, 1963)so a retrospective examination of necropsy materialbefore 1957 was undertaken in an attempt to resolvethis issue.

Materials and Methods

NECROPSIESThe necropsy records from 1949 to 1957 inclusive'Present address:Department of Virology, University of StellenboschMedical School, Private Bag 4, Tiervlei, CapeReceived for publication 9 June 1971.

were consulted and the histology was reviewed in allcases in which any of the following pathologicalchanges were recorded: liver necrosis (includingmiliary tubercles), adrenal haemorrhage or necrosis,digestive tract ulceration (non-peptic), splenicinfarction, atypical forms of pneumonia, and severemalnutrition. The tissue sections of 207- of a total of5,051 necropsies were re-examined and disseminatedherpetic infection was diagnosed in six cases. Livernecrosis, which was considered to be non-herpeticin origin, was noted in 82 other cases, and five ofthese were used as controls for the six herpetic casesin carrying out electron microscopy on the liverlesions.

EXPERIMENTALLY INFECTED WHITE MICESuckling and 3-week-old white mice (local highlyinbred strain) were inoculated intraperitoneally with800-1000 TCID50 of Herpes virus hominis strain E47liver which had been passed four times in culturesof vervet monkey kidney cells (Cercopithecusaethiops pygerythrus). All the mice died and showedfocal liver necroses. Portions of the livers of fourof the mice and of one control uninoculated mousewere fixed in 10% formol-saline and embedded inparaffin wax by routine methods; other portionswere fixed in alcohol and embedded in paraffin waxfor immunofluorescent studies (Sainte-Marie, 1962).The herpetic aetiology of the focal necroses wasconfirmed by the characteristic histological features,

79

on 6 May 2019 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.25.1.79 on 1 January 1972. Dow

nloaded from

http://jcp.bmj.com/

A. G. Rose and W. B. Becker

by specific immunofluorescence, and by the demon-stration of herpesvirus particles in the liver lesionsby electron microscopy.

ELECTRON MICROSCOPY

Several methods have been described for theexamination of wax-embedded tissues by electronmicroscopy (Zagury, Zeitoun, and Viette, 1966;Descarries, 1967; Morecki and Becker, 1968;Zagury, Pappas, and Marcus, 1968). The methodused was as follows: paired sections were cut fromthe paraffin blocks, one 5-8,u thick for routinehaematoxylin and eosin staining and the other20-80u thick. A lesion with suitable landmarks was

selected on the stained section so that the corre-

sponding area on the unstained thick section couldbe identified and removed with fine dissecting needlesand processed for electron microscopy. As an

alternative to the thick sections, the selected areas

were in some cases removed directly from the blockface if suitable landmarks were present.The selected fragments were dewaxed in warm

xylol, hydrated through graded ethanol solutions,fixed in osmic acid in veronal buffer pH 7 2 for one

hour at 4°C, dehydrated through graded ethanolsolutions and embedded in epon. Ultrathin sectionswere stained with lead citrate (Reynolds, 1963',followed by 50% uranyl acetate in distilled water,and examined in a Siemens Elmiskop 1 A at 80 Kvand magnifications of 20,000 or 40,000.

Results

CASE REPORTS

Six cases of disseminated herpesvirus infection were

discovered (Table I). The first two cases, one a whiteneonate, the second a malnourished 8-month-oldcoloured child, were examples of children dyingfrom florid disseminated herpetic infection withextensive liver necrosis as described in previouspublications (McKenzie et al, 1959; Becker et al,1963; Becker, Kipps, and McKenzie, 1968). In theremaining four cases the herpetic infection was notthe dominant feature. In addition to the herpeticnecroses in the liver, there were also portal pyaemiclesions in the liver in case 4 and miliary tubercles incase 6. All but the neonate were malnourished.

Case IA 12-day-old white female died from an illnesscharacterized by pyrexia, jaundice, diarrhoea, andattacks of apnoeic cyanosis. Hepatomegaly was

noted two days before death. Septicaemia was

diagnosed. The morbid anatomical and histologicalfindings (Fig. 1) of florid disseminated Herpesvirushominis infection were striking, but were not recog-nized as such. Death was due to the herpesvirusinfection.

Case 2An 8-month-old male coloured infant was admitted

CaselPost- Race, Sex, Malnutri- Days in Herpetic Licer Herpesvirus Other Relev,antmortenm Age tion Hospital Ulceration of Particles DataNo. Mucous Macroscopicallv Microscopically, Perc entage in Liver by

Membranes oJ Liuer ElectronA ecrosis Microscopy

1 E.F. 12 - 12 Oesophagus Classical florid disseminated 85289/1951 days herpesvirus infection2 C.M. 8 1-I Mouth and Classical florid disseminated 60 -r22/1954 months nose herpesvirus infection

Fatty liver3 B.M. 5 - 27 Not recorded Yellow foci Herpetic lesions, 50 Septic thrombo-42/1954 months sparse inclusions, plebitis, lung

mild cellular abscess andresponse empyema

4 B.F. 12 25 Not recorded Yellow foci Portal pyaemia, 10 Pneumonia,43/1954 months herpetic lesions septicaemia,

with mild cellular Mantouxresponse negative

5 C.M. 18 22 Mouth, Yelloss-grey Herpetic lesions, 10 Pulmonary148/1954 months tongue, foci sparse inclusions, tuberculosis,

pharynx, mild cellu!ar Mantouxoesophagus response negative

6 C.F. 6 7 Mouth, Many tubercles Tubercles but 10 - Pulmonary and20811957 months tongue, also herpetic miliary tuber-

oesophagus, lesions with culosis, Mantouxpharynx, sparse inclusions, negativelarynx good cellular

response

Table I Data on six children with disseminated Herpesvirus hominis infection

80



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

Disseminated Herpesvirus hominis (Herpes simplex) infection

'us-ear A--w

* nr~~~~~~~~~~~~~nLi. 41a"'4V"%~4Is%44.7gt*'ens' 'C' -~~~~~~~~vi

%.yVArt~~~~~~~~~c

as't'%'.

r~~~~~~~~~~~~~~~~~~~~~~~~~l

4C ~~~~~.4 .

411: 4s

;~~~wk~ ~ ~ 4ei:. wjm 4ci..

*- it ,t ¾'':S; Xt4

btte ** e~ *

*. r*.IS~~

B

'VA

A

*Ia_~:.A A focus of herpetic necrosis (haematoxylin andeosin, x 75).

B Edge of the area of necrosis. Numerous herpeticintranuclear inclusion bodies were seen (haematoxylinand eosin, x 320).C Liver cells showing intranuclear inclusion bodies(haematoxylin and eosin, x 1,200).

Fig. 1 Case 1: histology of the liver fromn the neonate.

C'

81

Alffiffiftmdmh.W:

4 p on 6 May 2019 by guest. P


j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

4 9- 4 t a :e:

- 4p-. 444 ,

I

~

.4 * 4. 1 .. *

1 ~~ ~ ~ ~t No ,44 t.4f , -.*

* V r 4l 4~ 4e4l 4.r;-0

* *.>#i.4Ot

* i;': .c, '4:s...:;;k,V/

44 4 t1*,8 ^ + J

*~ ~~~~~~~~~~~~~~~~- F

.g .4.,

4 .:444 44 .4 4~~~~~~~~

4* .*...4

* ~~~~4 ' 4.

..... .~~~44,

4,.

F.,

r

I

*4 ..**.e *

44,t' 8

.-4 I.4

A The focuis of herpetic necrosis in the very fatty liver is conspicuous by the dilated, blood-filled sinusoids(haematoxylin and eosin, -. 50).

Fig. 2 Case 2: histology of the liver.

...

*i ..t w

pa-~~~~~~~~~~~~~~~~~~~~~~~~~~.wq3wt

B Herpeticinclusions

(arrowed) withinthe nuclei ofliver cells at theperiphery of thearea of necrosis(haematoxylinand eosin,x 480).

.0a ...A

r...

0.

...:4

.iiA .; ... .: i- pq:

0

*4.& .;

wr

82 A. G. Rose and W. B. Becker

4

t %

-I

:

a*

4 , 4I

..1.0

*. p -:-

w. 4 .!,

t.6 .;

2Jil # '

V.t.

ik



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


because of weight loss, diarrhoea, oedema, andulceration of the mouth and nose. Examinationshowed a very ill, pale, malnourished child withpurpura and hepatomegaly. The child was moribundand died on the day of admission. The morbidanatomical findings described in the necropsy reportwere, in retrospect, those of florid disseminatedHerpesvirus hominis infection, which was the ap-parent cause of death. Review of the liver sectionsshowed the characteristic herpetic lesions (Fig. 2).The aetiology of the striking morbid anatomical andhistological findings was not appreciated at the timeof necropsy.

Case 3This 5-month-old Bantu male infant was admitted tohospital in coma with gastroenteritis. Two weeksafter admission a right-sided pneumonia andempyema were diagnosed. The child died on the27th day after admission. Necropsy showed septiccut-down wounds on the limbs and a right-sidedlung abscess and empyema, which was consideredto be the cause of death. The liver was fatty andmany rounded yellow foci were noted on the sub-capsular and cut surfaces. These areas of necrosiswere regarded as non-specific. The present reviewof the histological sections of the liver showed thatthe focal necroses were herpetic in origin. Herpeticintranuclear inclusions were present, though sparse,and there was some histiocytic cellular reaction.

Case 4A 12-month-old Bantu female presented withdiarrhoea and pneumonia. After two weeks' treat-ment she was apyrexial with a normal chest radio-graph. Ten days later her condition deteriorated andshe died a few days after. Necropsyrevealed broncho-pneumoia, ulceration of the terminal ileum, andnumerous tiny yellow foci scattered throughout theliver. These foci were originally interpreted as focalnecroses due to septicaemia. Review of the histologyshowed herpetic inclusions in cells bordering theareas of necrosis (Fig. 3). In addition to the herpeticfocal necroses, microscopy of the liver showed afew areas of congestion with masses of bacteria andinflammatory cells, probably the result of portalpyaemia.

Case SThis 18-month-old Coloured male was admitted tohospital suffering from diarrhoea, vomiting, andconjunctivitis. A discharge from the left ear was alsonoted. Necropsy revealed ulceration of the tongue,pharynx, and oesophagus. Death resulted frompulmonary tuberculosis with cavitation and sub-acute miliary tuberculosis of the spleen was present.

..4f' 9, 94r' ',i ., '

I^v 15 -*g*~ "is4~J$

'1~~~~~~~~~01A.

4.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

4 * * ICe q. * *

(haematoxylin.n A,

10~ ~ ~ ~ ~ *

ik U4)o

Of *5f

-#I

A Low-power view of an area of herpetic necrosis(haematoxylin and eosin, x SO).

t..v

v ;W440 ¢

B Herpetic inclusion bodies in cells bordering thearea of necrosis in the liver (haematoxylin and eosin,x 800).Fig. 3 Case 4: microscopy of the liver.

83

Awl. qt0-AIML .,

.*_fib

4



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


'4~ ~~ ~~~~~~~ ~"'

::'!*4

44.

J4

4f

'4,

A Mliliary tubercie with typical Langhans giant cells. The liver is fatty (haematoxylin and eosin, 160).

B In addition to the tubercles there were areas of herpetic necrosis. The herpetic lesion is surrounded by a zone of

inflammatory cells, which is regarded as an indication of healing (haematoxylin and eosin, '- 30).

SO

C Herpetic

intranuclear

inclusions (arro wed)

Aat the edge of the

area of necrosis

illustrated in Figure

4B (haematoxylin

and eosin, x 480).

Fig. 4 Case 6:

microscopy of the

liver.

84

.0



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


The necropsy report mentioned small, yellow-grey,rounded foci in the liver. Histologically these fociof necrosis were regarded as non-tuberculous buttheir herpetic nature was not recognized.

Case 6A Coloured female aged 6 months developedstomatitis associated with pyrexia and died after aweek in hospital. Necropsy showed emaciation,anaemia, stomatitis, pharyngitis, laryngitis, andoesophagitis. Death was due to a bilateral caseatingtuberculous bronchopneumonia and miliary tubercleswere present in liver and spleen. Microscopically, inaddition to the miliary tubercles in the liver, therewere sparse areas of focal necrosis which were notregarded as tuberculous. When the sections werereviewed, inclusions typical of a herpesvirus infectionwere seen in these areas of necrosis (Fig. 4).

ELECTRON MICROSCOPYHerpesvirus particles were demonstrated in the liverlesions of all six cases by electron microscopy (Figs.5 and 6). Similar particles were present in the livers

of the mice infected with Herpesvirus hominis (Fig.7). No viral particles were found in the livers of thefive control necropsy cases, nor of the control un-infected mouse. Viral particles were demonstratedby electron microscopy with less difficulty in caseswhere the viral inclusions were prominent than inthose where inclusions were scanty.

IMMUNOFLUORESCENCE STUDIESPositive specific immunofluorescence was noted inareas of herpetic liver necrosis in the four miceinoculated with Herpesvirus hominis, but not in theliver of the control uninoculated mouse. However,specific immunofluorescence could not be demon-strated in the herpetic lesions of the livers of the sixnecropsy cases. The latter result was not unexpectedsince in our experience (unpublished observations)specific immunofluorescence of herpesvirus antigenin tissues is readily rendered undemonstrable if, aswas the case, routine processing methods with pro-longed fixation in 10% formol-saline are used;however if the fixation time in formol-saline is keptto the minimum required for adequate fixation,

Fig. 5 Electron micrograph ofan ultrathin section of liver from case 2. Naked herpesvirus capsids are presentin the cell nucleus, x 60,000.

85



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


Fig. 6 Electronmicrograph of

an ultrathinsection of liverfrom case 5. Anenvelopedherpesvirusparticle is seenin the cytoplasmadjacent to thenuclearmembrane,x 60,000.

Fig. 7 Electronmicrograph ofanultrathin sectionof the liver of aninfected mouse.Nakedherpesviruscapsids arepresent in thecell nucleus andan envelopedparticle is in thecytoplasmadjacent to thenuclearmembrane,x 60,000.

herpesvirus antigen may be demonstrated.Templeton (1970) achieved success in four of hissix cases.

Discussion

Since the report by Hass (1935) many cases ofvisceral involvement in disseminated Herpesvirushominis infection have been reported from variousparts of the world and the subject is dealt with intextbooks (eg, Bedson, Downie, MacCallum, andStuart-Harris, 1967).

This study has shown that classical florid cases ofdisseminated Herpesvirus hominis infection occurredboth in the neonatal form and in association withmalnutrition in older children in Cape Town before1957 but that the diagnosis had been missed. Therecognition of the disease in 1957 in Cape Town wastherefore due to a new awareness of the disease andnot to the emergence of a more virulent strain ofherpesvirus. In establishing the diagnosis of dis-seminated Herpesvirus hominis infection, it wasnecessary to differentiate the disease from infectionwith other members of the herpesvirus group,namely cytomegalovirus and varicella virus. Thehistological distinction between lesions producedby cytomegalovirus and Herpesvirus hominis is notdifficult. In the absence of virus investigations, dis-seminated Herpesvirus hominis and varicella virusinfections can usually be differentiated with con-fidence by considering as a whole the epidemio-logical, clinical, morbid anatomical, and histologicaldata. These aspects of Herpesvirus hominis infectionhave been considered elsewhere (Becker et al, 1968).The features of disseminated varicella have beendescribed by a number of authors such as Johnson(1940); Cheatham,Weller, Dalan, andDower(1956);Nicolaides (1957); Rotter and Collins (1961), andTriebwasser Harris, Bryant, and Rhoades (1967).In particular, the absence of a skin rash inthe patients in the present series is a strongargument against varicella. In addition, col-lateral epidemiological evidence of varicellawould be expected, especially in those cases witha prolonged stay in hospital. Ideally confirmationof the aetiology can be obtained by specific immuno-fluorescence if the preparation and storage con-ditions of the tissues have been satisfactory. Electronmicroscopic examination ofwax-embedded tissues isa feasible way of confirming the diagnosis of in-fection with a virus of the herpes group even if thetissues have been stored for long periods. In thepresent series the blocks had been prepared as longas 18 years previously.

Several features of these cases are worthy ofcomment. In cases 3, 4, and 5 the length of stay inhospital relative to the known incubation period of

86

AL



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/


herpesvirus infection suggests that the infection wasacquired in hospital at a time when the potentialdanger of cross-infection with herpesvirus was notappreciated. This view was supported by the factthat cases 3 and 4 were in the same ward at the sametime, and were in fact consecutive necropsies per-formed on the same day. The neonate lived her entirelife of 12 days in hospital and one may only speculateas to whether she acquired her herpetic infectionfrom her mother or from some other source in thehospital. In the series reported by Templeton (1970),case 1 apparently contracted the herpesvirus infec-tion while in hospital where he was for over threeweeks. Of interest in cases 5 and 6 were the negativetuberculin tests in association with the miliarytuberculosis which suggests a state of anergy withdiminished cellular resistance and presumably de-creased resistance to the herpesvirus infection. Case3 of Templeton's (1970) series was also associatedwith extensive tuberculosis. A good histiocyticcellular response in the liver lesions in case 6 andto a lesser extent in cases 3, 4, and 5 was a note-worthy feature. In these patients death was notdirectly attributable to the herpetic infection (TableI), and the herpetic lesion is seen at a later stage ofits evolution than in cases 1 and 2 in which deathoccurred during the florid stage of the disease, andin which no cellular reaction was seen (Becker et al,1968).Following on the reports from Cape Town, similar

cases of disseminated Herpesvirus hominis infectionin older children in association with malnutrition ormeasles were reported from university centres inWest Africa (Armengaud, Baylet, Camain, Quenum,Guerin, and Schluep, 1963) and East Africa(Templeton, 1970). These reports strengthen ourbelief that disseminated Herpesvirus hominis infec-tion has 'always' occurred and is more common thanpublished reports suggest.

We wish to thank Dr J. G. Burger, Superintendent,Groote Schuur Hospital, for permission to publish;Professor J. G. Thomson for access to necropsymaterial and helpful criticism; Professor A. Kippsfor helpful criticism; Mrs P. McDowell and Miss K.Larsson for technical assistance.

References

Armengaud, M., Baylet, R. J., Camain, R., Quenum, C., Guerin. M.,and Schluep, R. (1963). Note pr6liminaire a l'etude de laprimo-infection herpetique de l'enfant africain (a propos de244 observations). Bull. Soc. med. Afr. noire Langue franC., 8,358-375.

Becker, W. B. (1966). The epidemiology of herpesvirus infection inthree racial communities in Cape Town. S. Afr. med. J., 40,109-111.

Becker, W. B., Kipps, A., and McKenzie, D. (1968). DisseminatedHerpes simplex virus infection: Its pathogenesis based onvirological and pathological studies in 33 cases. Amer. J. Dis.Child., 115, 1-8.

Becker, W., Naude, Du T. W., Kipps, A., and McKenzie, D. (1963).Virus studies in disseminated Herpes simplex infections: asso-ciation with malnutrition in children. S. Afr. med. J., 37, 74-76.

Bedson, S., Downie, A. W., MacCallum, F. O., and Stuart-Harris,C. H. (1967). Virus and Rickettsial Diseases of Man, 4th ed.,p. 125 et seq. Arnold, London.

Cheatham, W. J.. Weller, T. H., Dalan, T. F., Jr., and Dower, J. C.(1956). Varicella: A report of two fatal cases with necropsy,virus isolation and serologic studies. Amer. J. Path., 32, 1015-1035.

Descarries, L. (1967). An improved method for preparing cultivatednervous tissue for electron microscopic study. J. de Micro-scopie, 6, 313-320.

Hass, G. M. (1935). Hepato-adrenal necrosis with intranuclear in-clusion bodies. Amer. J. Path., 11, 127-142.

Johnson, H. N. (1940). Visceral lesions associated with varicella.Arch. Path., 30, 292-307.

McKenzie, D., (1961). Disseminated Herpes simplex infection. S. Afr.med. J., 35, 133-135.

McKenzie, D. Hansen, J. D. L., and Becker, W. (1959). Herpessimplex virus infection: Dissemination in Association withmalnutrition. Arch. Dis. Childh., 34, 250-256.

Morecki, R., and Becker, N. H. (1968). Human herpesvirus infection-its fine structure identification in paraffin-embedded tissue.Arch. Path., 86, 292-296.

Nicolaides, N. J. (1957). Fatal systemic varicella: -a report of threecases. Med. J. Aust., 2, 88-91.

Reynolds, E. S. (1963). The use of lead citrate at high pH as anelectron-opaque stain in electron microscopy. J. Cell Blol., 17,208-212.

Rotter, R., and Collins, J. D. (1961). Fatal disseminated varicella inadults: report of a case and a review of the literature. Wis.med. J., 60, 325-332.

Sainte-Marie, G. (1962). A paraffin embedding technique for studiesemploying immunofluorescence. J. Histochem. Cytochem., 10,250-256.

Templeton, A. C. (1970). Generalised herpes simplex in malnourishedchildren. J. clin. Path., 23, 24-30.

Triebwasser, J. H., Harris, R. E., Bryant, R. E., and Rhoades, E. R.(1967). Varicella pneumonia in adults. Report of seven casesand a review of literature. Medicine (Baltimore), 46, 409-423.

Zagury, D., Pappas, G. D., and Marcus, P. I. (1968). Preparation ofcell monolayers for combined light and electron microscopy:staining in blocks. J. Microsc., 7, 287-292.

Zagury, D., Zeitoun, P., and Viette, M. (1966). M6thode d'6tude decultures de cellules par la Microscopie 6lectronique: la Culturesur support d'6pon. C.R. Acad. Sci. [D] (Paris), 262,1458-1459.

87



j.com/

J Clin P


nloaded from

http://jcp.bmj.com/

Documents

Disseminated Herpesvirus (Herpes simplex) · 2 C.M. 8 1-I Mouth and Classical florid disseminated 60-r 22/1954 months nose herpesvirus infection Fatty liver 3 B.M. 5-27 Notrecorded