Journal of Wildlife Diseases, 36(1), 2000, pp. 1-12 
? Wildlife Disease Association 2000 
CLINICAL AND PATHOLOGICAL FINDINGS OF A NEWLY 
RECOGNIZED DISEASE OF ELEPHANTS CAUSED BY 
ENDOTHELIOTROPIC HERPESVIRUSES 
Laura K. Richman,^^^ Richard J. Montali,^ Richard C. Cambre,^ Dennis Schmitt,^ Dougias 
Hardy,^ Thomas Hiidbrandt," Roy G. Bengis,^ Fayez M. Hamzeh,^ Al<bar Shahlcoiahi,^ and 
Gary S. Hayward^ 
' Smithsonian, National Zoological Park, Washington, DC 20008, USA 
^ Southwest Missouri State University, Springfield, Missouri 65804, USA 
3 Dickerson Park Zoo, Springfield, Missouri 65803, USA 
" Institute for Zoo Biology and Wildlife Research, Berlin, Germany 
^ Kruger National Park, Skukuza 1350, Republic of South Africa 
''Johns Hopkins School of Medicine, Baltimore, Maryland 21205, USA 
'Corresponding author (e-mail: lkrichma@jhmi.edu) 
ABSTRACT: The unique clinical and pathological findings in nine Asian (Elephas maximus) and 
two African (Loxodonta africana) elephants from North American Zoos with a highly fatal disease 
caused by novel endotheliotropic herpesviruses are described. Identification of the viruses by 
molecular techniques and some epidemiol?gica! aspects of the disease were previously reported. 
Consensus primer polymerase chain reaction (PCR) combined with sequencing yielded molecular 
evidence that confirmed the presence of two novel but related herpesviruses associated with the 
disease, one in Asian elephants and the second in African elephants. Disease onset was acute, 
with lethargy, edema of the head and thoracic limbs, oral ulc?ration and cyanosis of the tongue 
followed by death of most animals in 1 to 7 days. Pertinent laboratory findings in two of three 
clinically evaluated animals included lymphocytopenia and thrombocytopenia. Two affected young 
Asian elephants recovered after a 3 to 4 wk course of therapy with the anti-herpesvirus drug 
famciclovir. Necropsy findings in the fatal cases included pericardial effusion and extensive pe- 
techial hemorrhages in the heart and throughout the peritoneal cavity, hepatomegaly, cyanosis of 
the tongue, intestinal hemorrhage, and ulc?ration. Histologically, there were extensive microhem- 
orrhages and edema throughout the myocardium and mild, subacute myocarditis. Similar hem- 
orrhagic lesions with inflammation were evident in the tongue, liver, and large intestine. Lesions 
in these target organs were accompanied by amphophilic to basophilic intranuclear viral inclusion 
bodies in capillary endothelial cells. Transmission electron microscopy of the endothelial inclusion 
bodies revealed 80 to 92 nm diameter viral capsids consistent with herpesvirus morphology. The 
short course of the herpesvirus infections, with sudden deaths in all but the two surviving ele- 
phants, was ascribed to acute cardiac failure attributed to herpesvirus-induced capillary injury 
with extensive myocardial hemorrhage and edema. 
Key tvords:     Elephant, Elephas maximus,  endotheliotropic viruses, famciclovir, herpesvirus, 
Loxodonta africana, new disease, penciclovir. 
INTRODUCTION ings of a highly fatal disease of elephants 
in North America caused by two related 
Elephants are exceptionally charismatic ^^^ ?j^y^^^^ ^^^^j endotheHotropic her- 
megavertebrates; the Asian (Elephas max- pesviruses (Richman et al., 1999a). In that 
imus) and African (Loxodonta africana) ^^^^^ sequences from tv^^o herpesvirus 
species are hsted respectively as endan- gg^g regions, terminase and DNA poly- 
gered and threatened by The World Con- merase, showed 80% and 65% DNA ho- 
servation Union (lUCN). To increase ele- mology, respectively between the viruses 
phant numbers, a great deal of effort is that are lethal for Asian and African ele- 
underway by zoo organizations to breed phants. A similar disease attributed to an 
these majestic animals naturally and by as- endotheliotropic herpesvirus has been re- 
sisted insemination techniques. However, ported in a young Asian elephant from a 
too many young elephants still die from circus in Switzerland (Ossent et al., 1990) 
unknown causes. Recently, we reported and was shown to have the same viral se- 
the virological and epidemiological find- quences as in our cases (Richman et al.. 
JOURNAL OF WILDLIFE DISEASES, VOL. 36, NO. 1, JANUARY 2000 
1999a). In our series, a total of 11 cases of 
endotheliotropic herpesvirus infections 
were identified from eight zoos in North 
America by polymerase chain reaction 
(PCR). Six of the cases were found retro- 
spectively after identification of the index 
case involving a 16-mo-old Asian elephant 
at the National Zoological Park in 1995 
(Richman et al., 1999a). The disease was 
fatal in six young Asian elephants repre- 
senting 18% of that species born in cap- 
tivity from 1983 to 1996 (Richman et al., 
1999a). One adult Asian elephant and two 
African elephants (a calf and an adult) also 
died, and two young Asian elephants re- 
cently survived the disease after treatment 
with an anti-herpesvirus drug, famciclovir. 
The relationship between the Asian and 
African elephant endotheliotropic herpes- 
viruses is under further investigation; how- 
ever, they appear to be closely related, 
novel herpesviruses that produce nearly 
identical lesions in the two elephant spe- 
cies (Richman et al., 1999a). Additionally, 
localized African elephant skin and vulval 
lesions were found to contain viral se- 
quences nearly identical to the virus that 
caused the highly fatal infections in Asian 
elephants (Richman et al., 1999a). 
The purpose of this paper is to report 
the clinical and pathological findings of 
this new herpesvirus disease of elephants 
and to describe a unique pattern of lesions 
in the organs targeted which is attributed 
to the endotheliotropism of the virus. In 
addition, we have developed an early di- 
agnostic method for this disease by a 
whole blood PCR test, and propose a pos- 
sible source of the herpesvirus that caused 
the deaths of the two African elephants 
with endotheliotropic disease. The gross 
and microscopic findings together with the 
molecular virological data and diagnostic 
tests support these herpesviruses as the 
causative agents of this newly recognized 
disease syndrome. 
MATERIALS AND METHODS 
General 
Clinical and/or postmortem samples were 
obtained from 11 elephants from North Amer- 
ican zoos with endotheliotropic herpesvirus in- 
fection which included seven Asian and two Af- 
rican elephants that died, and two Asian ele- 
phants that survived. A detailed account of the 
clinical management of one of the latter ele- 
phants is published elsewhere (Schmitt and 
Hardy, 1998). 
Clinical and patliological evaluations of recent 
cases (1995-98). 
Signalment and disease status of elephant 
(E) #1 through E#4 were assessed (Table 1); 
clinical evaluations were dependent upon the 
tractability of the animals at the time of illness 
but usually included physical examinations, col- 
lection of one or more blood samples (EDTA 
and heparinized blood) for hemograms, blood 
bacterial and viral cultures, and serum for se- 
rological testing, vitamin E levels (E. Dieren- 
feld. Wildlife Conservation Society, New York 
City, New York, USA), and to assay famciclovir 
serum levels in E#3. Elephant #3 and E#4 
were treated with an anti-herpesvirus drug, 
famciclovir (Famvir, SmithKline Beecham, 
Philadelphia, Pennsylvania, USA) administered 
orally or rectally at 500 mg/70 kg, three times 
a day for 3 to 4 wk (Schmitt and Hardy, 1998). 
Complete necropsies were performed on 
E#l and E#2 and representative samples were 
taken from all organs and fixed in buffered 10% 
formalin. All organs and tissues were embed- 
ded in paraffin, sectioned at 5 microns thick- 
ness, stained with hematoxylin and eosin, and 
examined by light microscopy. Additional sec- 
tions of selected formalin-fixed heart, liver, 
tongue and intestinal tract were post-fixed in 
2.5% gluteraldehyde and 2% osmium tetroxide, 
dehydrated in a series of graded alcohols, and 
embedded in epoxy resin (Steele et al., 1997). 
Semi-thin sections were stained with toluidine 
blue, and ultra-thin sections were stained with 
uranyl acetate and lead citrate for electron mi- 
croscopy (Steele et al., 1997). From E#l, fresh 
liver and spleen and paired serum samples 
were submitted for fluorescent antibody tests 
(FAT) for leptospirosis (Leptospira interrogans 
serotypes pomona, hardjo, icterohaemorrhagia/ 
copenhageni, grippotyphosa, canicola, bratis- 
lava, and sejroe; New York State College of 
Veterinary Medicine, Ithaca, New York, USA) 
and fresh heart for encephalomyocarditis virus 
(EMCV) cultures (J. Gaskin, University of Flor- 
ida, Gainesville, Florida, USA). Replicate sec- 
tions of heart, liver, tongue, lymph nodes, 
spleen, kidney, and intestinal tract were frozen 
at ? 70 C for further viral isolation attempts. 
Retrospective pathological studies 
A search for pathology material from poten- 
tial cases of this herpesvirus disease was carried 
RICHMAN ET AL.?A NEW DISEASE OF ELEPHANTS CAUSED BY ENDOTHELIOTROPIC HERPESVIRUSES 
out by sending a survey to all participating zoos 
via the American Zoo and Aquarium Associa- 
tion (AZA, Silver Spring, Maryland, USA) Spe- 
cies Survival Plan (SSP, Portland, Oregon, USA 
and Indianapolis, Indiana, USA) and by review- 
ing the Asian and African elephant studbook 
mortality records (held by the AZA). There 
were no reports of herpesvirus infections as- 
sociated with elephant deaths from the survey 
or recorded in the studbooks from early in the 
century in North America. Pathology records, 
microscopic slides and tissues (frozen and 
fixed) were obtained from over 20 elephants 
whose clinical and pathological descriptions 
had some similarities with E#I (the index case) 
and the case reported in an Asian elephant in 
Switzerland (Ossent et al., 1990). Additionally, 
archival paraffin blocks of lung tissue reported 
to have herpesvirus inclusion bodies from 
healthy wild African elephants (McCuUy et al., 
1971) (received from N. Kriek, Onderstepoort 
Veterinary College, Pretoria, Republic of South 
Africa) were obtained. The fixed tissues were 
processed as described for E#l and E#2 for 
light and electron microscopy; all slides re- 
ceived and any new slides prepared were re- 
viewed, all major organs and tissues were ex- 
amined by light microscopy, and the data col- 
lated. 
PCR assay and sequencing 
Polymerase chain reaction was performed on 
DNA extracted from either fresh or frozen (?80 
C) peripheral blood from two ill elephants (E#3 
and E#4) with clinical signs of endotheliotropic 
herpesvirus disease, and on DNA isolated from 
lung tissue from three healthy, wild African ele- 
phants containing lymphoid nodules reported to 
include herpesvirus inclusions (McCuUy et al., 
1971) and processed for PCR as described pre- 
viously (Richman et al., 1999a). Specific single 
primer sets for the elephant herpesvirus termi- 
nase and DNA polymerase gene regions were 
constructed from the sequence initially obtained 
from cases E#l and E#2: Terminase primers for 
Asian elephants (5'GTACGTCCTTTCTAGCT- 
CAC 3' and 5'GTGTCGGCTAAATGTTCTTG 
3'); terminase primers for African elephants 
(5'AATGTGATATCCTACGTATG 3' and 5' 
GTACTATATCTTATCATGTC 3'); and DNA 
polymerase primers for both elephant species (5' 
GTGTCTGGCTATAGCAGAGT 3' and 5'CA- 
CATCGATACGGAATCTCT 3'). 
Polymerase chain reaction was performed 
with these primers in a 50 jj,l reaction volume 
containing PCR SuperMix (Gibco/RRL-Life 
Technologies, Gaithersburg, Maryland, USA), 
0.3% (v/v) glycerol, and 20 pmol of each prim- 
er.  Polymerase chain reaction was completed 
using the following protocol: Denaturation at 
94 C for 1 min, annealing at 50 C for 1 min, 
and extension at 72 C for 2 min for a total of 
36 cycles. The final extension was performed at 
72 C for 7 min. The PCR product was visual- 
ized on a 1.5% agarose gel stained with ethi- 
dium bromide. 
Serological testing 
Selected serological tests were performed via 
the National Animal Disease Laboratories 
(NADL; Ames, Iowa, USA) on sera from three 
adult elephants (two Asian, one African) from 
the National Zoological Park (Washington DC, 
USA) as well as banked frozen sera from herd- 
mates from two other zoos (located in Rronx, 
New York, USA and Springfield, Missouri, 
USA). These were the sites where previous 
deaths from the endotheliotropic herpesvirus 
had occurred. Serologie tests for antibodies to 
the following were performed: bluetongue vi- 
rus by enzyme-linked immunosorbent assay, 
epizootic hemorrhagic disease virus of deer by 
agar gel immunodiffusion (AGID), bovine her- 
pesvirus 1 by serum neutralization (SN), bovine 
herpesvirus 2 by SN, bovine herpesvirus 4 by 
immunoperoxidase (IPT), malignant catarrhal 
fever virus by IPT, equine herpesviruses by SN, 
equine viral arteritis virus by SN, encephalo- 
myocarditis virus by SN, equine infectious ane- 
mia virus by AGID, equine adenovirus by IPT, 
pseudorabies virus by SN and porcine repro- 
ductive and respiratory syndrome virus by IPT. 
IHigh performance liquid chromatographiy (IHPLC) 
analysis 
Famciclovir (Famvir, SmithKline Reecham, 
Philadelphia, Pennsylvania, USA) is a pro-drug 
that is rapidly and completely converted to 
penciclovir by liver esterases (Ver? Hodge et 
al., 1989). Penciclovir (SmithKline Reecham, 
Philadelphia, Pennsylvania, USA) is the com- 
pound that circulates in the blood after admin- 
istration of famciclovir. We used, with minor 
modifications, a previously described method 
for the determination of penciclovir concentra- 
tions in serum (Fowles and Pierce, 1989) from 
elephant #3. The calibrators (range 10?10,000 
ng/niL), quality assurance samples, and E#3 se- 
rum samples were spiked with acyclovir (Sig- 
ma, St. Louis, Missouri, USA) (50 jj,L of 100 
(ji,g/mL) which was used as the internal stan- 
dard. Trichloroacetic acid (0.5 mL) was added 
to precipitate the proteins. The samples were 
centrifuged at 1,200 rpm in an Eppendorf 
235C centrifuge (Fisher Scientific, Pittsburgh, 
Pennsylvania, USA) for 2 min and the super- 
natants were transferred to a new tube and 
then dried in a Speed Vac (Savant Instrument 
JOURNAL OF WILDLIFE DISEASES, VOL. 36, NO. 1, JANUARY 2000 
TABLE 1.     Signalment and disease status of four elephants with endotheliotropic herpesvirus disease. 
ID Species Sex Age Location Onset Course Status 
Elephant #1 Asian F 16 mo Washington, DC Apr-95 5 days Died 
Elephant #2 African M 11 mo California Aug-96 3 days Died 
Elephant #3 Asian F 17 mo Missouri Nov-97 10 days Survived" 
Elephant #4 Asian M 16 mo Florida Sep-98 12 days Survived" 
^ Animals were treated with famcicloxdr, be, Tinning .3 to 5 days after onset. 
Inc., Holbrook, New York, USA). The samples 
?were resuspended in 300 |xL of 0.001 M 
NaH2P04. Aliquots of 100 jjuL were injected 
using an autoinjector (Waters Associates, Inc., 
Milford, Massachusetts, USA) onto a C-18 re- 
verse phase HPLC column (Waters Associates, 
Inc., Milford, Massachusetts, USA) and then 
were eluted with 0.01 M NaH2P04 (pH 7): 
methanol (9:1) at a flow rate of 1 mL/min using 
Waters HPLC pump. The signal was detected 
using Waters spectrophotometric detector at 
UV absorption of 254 nm. Data acquisition was 
done using Waters Millennium Chromatogra- 
phy Manager. Samples were obtained from E#3 
several times each day during the treatment pe- 
riod and were compared to banked serum sam- 
ples from the same elephant before treatment 
commenced. 
RESULTS 
Clinical findings 
From the clinical evaluations of E#l? 
E#4 (Table 1), the onset of the disease ap- 
FlGURE 1. Asian elephant #3 has subcutanous 
edema of the head, neck, and thoracic limbs attri- 
buted to endotheliotropic herpesvirus disease. 
peared to be sudden with a course of 1 to 
7 days. The initial clinical signs were leth- 
argy, anorexia, mild colic, edema of the 
head (especially the face and proboscis) 
neck and thoracic limbs (Fig. 1). E#l had 
a diffusely cyanotic tongue and oral ulcers 
on the day that she died (day 5). Cyanosis 
of the tongue in E#2 was not observed, 
but in E#3 cyanosis began at the tip of the 
tongue and moved caudally on the third 
and fourth day. E#4 had swelling and cy- 
anosis of the tongue noted 4 days after the 
onset of illness. Pertinent laboratory tests 
included hemograms early in the course of 
the disease in the three Asian elephants 
which showed mainly lymphopenia and 
thrombocytopenia in two of the elephants, 
one of which also was anemic (Table 2). 
E#l and E#2 died on the fifth and third 
day, respectively after the onset of the dis- 
ease. In E#3, the lingual cyanosis and head 
and neck edema resolved over the next 
week, approximately 9 days after begin- 
ning treatment with famciclovir. The ele- 
phant gained strength and resumed eating. 
Semiquantitative blood virus levels mea- 
sured daily from day 3 after disease onset 
decreased steadily to a non-detectable lev- 
el by 8 wk after the onset of illness. Serum 
levels of penciclovir rose to 863 ng/ml af- 
ter the initial dose and fluctuated between 
0 (8 hr post-treatment) and 4,365 ng/ml (1 
hr post-treatment) over the 3 wk course of 
oral famciclovir administration. 
Patliological findings (recent and retrospective 
cases). 
The necropsied elephants E#l and E#2 
and seven of the 20 elephants reviewed 
retrospectively that died between 1983 
and 1993 all had similar lesions attributed 
RICHMAN ET AL.?A NEW DISEASE OF ELEPHANTS CAUSED BY ENDOTHELIOTROPIC HERPESVIRUSES        5 
"3 
?a 
d 
13 
d 
&^ 
-d 
0 
"3 
0 
13 
d 
0) 
d 
?d 
d 
13 
13 
0 
0 
?. 
? ? 
o o 
CD 
O CD 
1?1 
X ? ^ X ? 
-* -* 
sj ^ 
~? "i ? o ? CO i-H c? lO i-H -1^ 
X ? CO X ? 
1?1 
CD (M 
? ? 
1 "1 
O O 
^ in ^ CD 
M lO q 
X CO c? X \6 
t~ CO ? 
"1 ^ "1 Q) 
o 0) o cti 
^ ClJ ^ :3 p 
^ X o ^ X 
^ 00 13 ni 
lO 
"d. ~? ? c; in ? M 1?1 00 11? i-H CO 
X CO oi X M 
^ (M 
CO CO 
"1 "1 
o O 
^ ID in ^ 00 
X -* X 
uo 
CO ? 
^ ^ 
? 
-3? 
"rt ? 
s S S JH rt 
fa "rt a? S 
0) ?J ? rt fa fa rt S 
OJ tD 
s M # CO * S -* * 
FIGURE 2. Left ventricular endocardia! surface of 
the heart from Asian elephant #1 showing subendo- 
cardial ecchymotic hemorrhages due to endotheh- 
otropic herpesvirus disease. 
to the endotheliotropic herpesviruses. Six 
of the retrospective cases were Asian ele- 
phants from zoos in New York, Oklahoma, 
Illinois, and Missouri in the USA, and in 
Ontario, Canada (2); one was an African 
elephant from a zoo in Texas (USA). 
Gross findings typically included peri- 
cardial effusion with extensive petechial to 
ecchymotic hemorrhages involving the 
epi- and endocardial heart surfaces (Fig. 
2) and throughout the myocardium. In ad- 
dition, diffusely scattered petechiae were 
within all of the visceral and parietal peri- 
toneal serous membranes and variably, 
there Nvas cyanosis of the tongue (Fig. 3). 
There \vas also hepatomegaly, and variably 
oral, laryngeal and large intestinal ulcers. 
Light and electron microscopic findings 
Pertinent microscopic findings consisted 
of extensive microhemorrhages through- 
out the heart (Fig. 4) and tongue associ- 
ated with edema, and mild infiltrates of 
lymphocytes, monocytes and a few neutro- 
phils between myofibers. There was mild 
to moderate hepatic sinusoidal expansion 
with multifocal subacute inflammation, 
and mild hepatocellular vacuolar degen- 
erative changes. Ulcers of the oral and la- 
6        JOURNAL OF WILDLIFE DISEASES, VOL. 36, NO. 1, JANUARY 2000 
FIGURE 3. Cross section of the tongue from 
Asian elephant #1 demonstrating diffuse sweHing and 
cyanosis attributed to endotheHotropic herpes\4rus 
disease. 
FIGURE 4. Photomicrograph of heart from Asian 
elephant #1, demonstrating extensive hemorrhage 
and separation of myocardial fibers attributed to en- 
dotheliotropic herpes\4rus. Bar = 150 fxm. 
FIGURE 5. Photomicrograph of liver from Asian 
elephant #1 that died \vith endotheliotropic herpes- 
\4rus disease, showing an intranuclear viral inclusion 
body within a sinusoidal endothelial cell (arrow). Bar 
= 25 fxm. 
ryngeal mucous membranes in E#l were 
acute, with necrotic surface cells still intact 
in some areas. The endothelial cells of cap- 
illaries in the myocardium (9/9 animals) 
and tongue muscle (from 6/6 animals from 
which tongue was available), and within 
the hepatic sinusoids of the liver (9/9 ani- 
mals) contained amphophilic to basophilic 
intranuclear viral inclusion bodies (Fig. 5). 
The endothelial cells with the viral inclu- 
sion bodies were in close association with 
the microhemorrhages in the heart and 
tongue. The inclusion bodies were less of- 
ten seen in capillary endothelial cells in 
the lamina propria and smooth muscle lay- 
ers of the intestinal tract, but were not ev- 
ident in any of the ulcers or in blood ves- 
sels larger than capillaries. Ultrastructural 
microscopic studies of the endothelial in- 
clusion bodies in all nine cases revealed 
80?92 nm particles morphologically con- 
sistent with herpesviruses (Figs. 6, 7). His- 
RICHMAN ET AL.?A NEW DISEASE OF ELEPHANTS CAUSED BY ENDOTHELIOTROPIC HERPESVIRUSES        7 
FIGURE 6. Transmission electron micrograph 
(TEM) of heart from African elephant #2 that died 
\vith endotheliotropic herpesvirus disease, demon- 
strating an intranuclear inclusion body within a cap- 
illary endothelial cell. Herpesvirus particles (arrows) 
are seen emerging from the inclusion body. Bar = 
500 nm. 
tologically, pulmonary nodules from archi- 
val African elephant tissues -were com- 
posed of multiple large lymphoid follicles 
(Fig. 8) which surrounded epithelial cells 
that contained intranuclear inclusion bod- 
ies (Fig. 9). These epithelial cells often 
formed syncytia. 
Clinical laboratory findings 
In E#l, aerobic and anaerobic bacterial 
cultures of heart blood, pericardial and 
peritoneal fluid, cerebrospinal fluid, axil- 
lary lymph node, and liver were negative, 
and bacterial cultures of multiple seg- 
ments of the gastrointestinal tract yielded 
no enteric pathogens. Co-cultivation of liv- 
er and heart were performed as previously 
described (Richman et al., 1999a) and also 
were negative for any of the known human 
or animal herpesviruses or any other vi- 
FlGURE 7. Higher magnification TEM of the in- 
tranuclear inclusion body from Figure 6, sho\ving 
greater detail of the emerging herpesvirus nucleocap- 
sids (arrows). Bar = 200 nm. 
ruses. In addition, serology and cultures of 
heart tissue were negative for encephalo- 
myocarditis virus; FAT and paired sera 
were negative for Leptospira interrogans 
antigens and antibodies. Blood alpha to- 
copherol level was 0.62 |ji,g/ml (reference 
range is 0.2 |xg/ml to 0.9 |jLg/ml for Asian 
elephants) at the time of her illness. 
PCR and viral sequencing 
Polymerase chain reaction, followed by 
sequencing of DNA extracted from pe- 
ripheral blood from E#3 and E#4 con- 
tained nearly identical endotheliotropic 
herpesvirus sequences in the terminase 
gene region (up to 3 base pairs difference 
from some of the deceased Asian ele- 
phants, but all are silent mutations) as the 
target tissues from the deceased Asian el- 
ephants. DNA extracted from pulmonary 
tissue of wild African elephants with mor- 
phological  evidence  of herpesvirus  con- 
8        JOURNAL OF WILDLIFE DISEASES, VOL. 36, NO. 1, JANUARY 2000 
FIGURE 8. Pulmonary lymphoid nodules from a 
healthy, free ranging adult African elephant. FoUic- 
ular lymphoid hyperplasia surrounds epithelial cells 
which contain intranuclear herpes\4ral inclusion bod- 
ies, shown in Figure 9. Bar = 250 fxm. 
tained viral sequences in the DNA poly- 
merase gene region that had 100% protein 
identity with the herpesvirus that was fatal 
for the two African elephants in our study 
group. 
Serological studies 
One herdmate of E#l had low serum 
antibody titer to bovine herpesvirus-4, and 
several others tested had low titers to ma- 
lignant catarrhal fever virus (alcelaphine 
herpesvirus 1) and to equine adenovirus. 
Serosurveys from remaining viruses listed 
were all negative. Herd mates from the 
Swiss case (Metzler et al., 1990) also had 
low titers considered cross reactive to bo- 
vine herpesviruses but the significance of 
these s?rologie findings is currently un- 
known since there was no consistent pat- 
tern, nor were herpesviruses isolated from 
any of the cases. 
FIGURE 9. Higher magnification of the lung nod- 
ule from Figure 8, demonstrating intranuclear her- 
pes\4ral inclusion bodies (arro\vs) \vithin epithelial 
syncytial cells. Herpes\4ral protein sequences from 
the DNA polymerase gene region obtained from 
these African elephant pulmonary nodules are iden- 
tical to those found in deceased African elephants 
\vith endotheliotropic herpesvirus disease. Bar = 25 
fxm. 
HPLC for penciclovir 
Serum levels of penciclovir from E#3 
varied from 97 ng/ml to 4,365 ng/ml, de- 
pending on the time interval between 
treatment and blood collection. Lower lev- 
els of serum penciclovir were present in 
the early morning, 8 hr after the last dose 
of famciclovir was administered. Higher 
values of serum penciclovir correlated with 
treatment administration 1 to 3 hr prior to 
blood collection (data not shown). These 
concentrations are comparable to penci- 
clovir concentrations in humans after ad- 
ministration of a 500 mg dose of famciclo- 
vir (Vere Hodge et al., 1989; Boike et al., 
1994). 
RICHMAN ET AL.?A NEW DISEASE OF ELEPHANTS CAUSED BY ENDOTHELIOTROPIC HERPESVIRUSES 
DISCUSSION 
The acute nature and high fatality rate 
in clinically affected elephants of this her- 
pesvirus disease is unusual for infections 
caused by herpesviruses in humans and 
other animals. One of the major factors 
considered important in the epidemiology 
of the elephant herpesviruses is the prem- 
ise of the herpesvirus being innocuous in 
one elephant species and highly pathogen- 
ic for the other. This relationship is known 
for a number of herpesviruses including 
herpesvirus B, latent in Macaca spp. and 
potentially fatal in humans, and malignant 
catarrhal fever herpesviruses carried by 
sheep or \vildebeest and producing overt 
disease in cattle and acelaphine antelopes 
(Schuller et al., 1990). The evidence indi- 
cating probable cross-species transmission 
of a herpesvirus from benign lesions in 
otherwise asymptomatic African elephants 
to susceptible Asian elephants was the 
finding of nearly identical herpesvirus se- 
quences in these two species (Richman et 
al., 1999a). This suggested that otherwise 
healthy African elephants are a source of 
the herpesvirus that is highly pathogenic 
for Asian elephants. 
Another important factor considered in 
the pathogenicity and the unusual nature 
of the lesions observed in these elephant 
herpesvirus infections is the endotheli- 
otropism of these viruses. Most animal 
herpesviruses are epitheliotropic or have a 
predilection for nervous tissue. The target 
organs for many of the disseminated her- 
pesvirus infections in mammals are usually 
the liver, adrenal glands and brain, in 
which the herpesvirus forms characteristic 
intranuclear inclusion bodies and produces 
foci of necrosis in the parenchymal cells. 
The finding of intranuclear inclusion bod- 
ies in the endothelial cells of capillaries in 
the elephant lesions provides histomor- 
phological evidence of their tropism for 
these vascular cells. In fact, these endo- 
thelial inclusion bodies found in all nine 
elephants that died of the herpesvirus in- 
fection are histological hallmarks and may 
be pathognomonic for this disease syn- 
drome. 
It has been reported (Metzler et al., 
1990) that elephants that had contact with 
the Swiss case had serological evidence of 
exposure to bovine herpesvirus 2 (BHV2). 
Because of this report, we tested conva- 
lescent serum samples from Asian ele- 
phant E#3 for exposure to or cross-reac- 
tivity with BHVl, BHV2 and BHV4 (data 
not shown). All convalescent serum sam- 
ples from E#3 were negative by ELISA for 
BHVl, BHV2 and BHV4 (R. Eberle, pers. 
comm.) and negative by immunofluores- 
cence for BHVl, BHV2, BHV4 and other 
ungulate herpesviruses including equine 
herpesviruses 1 and 4 (M. Kennedy, pers. 
comm.). This supports our molecular find- 
ings, and demonstrates that the elephant 
herpesviruses are unique in immunogenic- 
ity such that cross reactivity with other 
herpesviruses is not detected. Additionally, 
the vitamin E (alpha-tocopherol) level in 
E#l was within normal range for Asian el- 
ephants. In the earlier cases, at the time 
of their deaths, hypovitaminosis E was 
considered in the differential diagnosis of 
several of the elephants in our study group 
due to ecchymotic hemorrhages that were 
seen at necropsy. 
The endotheliotropism of the elephant 
herpesviruses also may explain the organ 
tropism of these viruses for the heart, liver 
and tongue, the capillary endothelial cells 
of which harbor the inclusion bodies. The 
reason for the specific tropism for endo- 
thelial cells mainly in these targeted organs 
is not clear, but may reflect some differ- 
ence in metabolic properties or surface re- 
ceptor characteristics between endothelial 
cells in different organs. Other instances 
of endotheliotropism for herpesviruses 
have been reported with varicella-zoster 
virus (VZV) in children with fatal dissem- 
inated infections, and more recently with 
human herpesvirus-8 (HHV8) and its as- 
sociation with Kaposi's sarcoma, which are 
thought to be derived from endothelial- 
like spindle cells (Moore and Chang, 
1995). 
10 JOURNAL OF WILDLIFE DISEASES, VOL. 36, NO. 1, JANUARY 2000 
Based on these morphologic findings, 
the proposed pathogenesis is as follows: 
once the elephant becomes viremic (pos- 
sibly via circulating lymphocytes after ex- 
posure to the virus), ensuing viral repli- 
cation in the heart leads to endothelial cell 
damage resulting in capillary leakage and 
severe intramyocardial hemorrhage and 
edema. The magnitude of this damage, as 
reflected by the microscopic lesions, leads 
to cardiac failure either by disruption of 
the conducting system, the mass effect of 
the intracardiac swelling, or myocardial is- 
chemia and localized metabolic distur- 
bances. Cardiac insufficiency may also 
contribute to glossal cyanosis. The regres- 
sion of lingual cyanosis as noted in E#3 
treated with famciclovir reflects a similar 
process of capillary injury and leakage 
within the tongue, followed by resolution 
as recovery progressed. The severe hem- 
orrhage and congestion was manifested as 
the blue discoloration noted clinically in 
E#3 and involving the entire tongue of el- 
ephant E#l in the terminal stages of the 
herpesvirus infection. 
In support of this pathogenesis, E#3 and 
E#4 were found to be viremic by PCR ear- 
ly in the disease, had all the classic clinical 
signs, and were treated with famciclovir 
for 3 to 4 wk. After several days of treat- 
ment, at a time when elephants would nor- 
mally die, appetite returned, lingual cya- 
nosis slowly regressed and edema was no 
longer visible. In E#3, temporal semiquan- 
titative PCR showed decreasing detectable 
blood viral load that coincided with serum 
famciclovir at therapeutic levels (data not 
shown) which strongly suggests a benefi- 
cial effect of the antiherpesvirus drug. The 
recovery of the second young elephant 
(E#4) that was treated with famciclovir 
further supports the efficacy of this drug 
in the treatment of these elephant herpes- 
virus infections. 
Prior to the identification and descrip- 
tion of this highly fatal herpesvirus disease 
in captive elephants (Ossent et al., 1990; 
Richman et al., 1999a), there existed only 
several reports of herpesviruses occurring 
in skin papillomas and pulmonary nodules 
of African elephants by light and electron 
microscopy, but no references to herpes- 
virus in Asian elephants. At the time, these 
lesions with herpetic inclusion bodies in 
the African elephants were considered in- 
cidental or localized findings with no sys- 
temic illnesses associated with them, and 
no documentation of herpesvirus isolation 
(McCuUy et al., 1971; Jacobson et al., 
1986). The agent observed in the African 
elephant cutaneous papillomas is homolo- 
gous with the virus lethal for Asian ele- 
phants in the terminase gene region (Rich- 
man et al., 1999a). In that study, the same 
virus was also identified by PCR in biop- 
sies of the lymphoid patches from the dis- 
tal vaginal tract (vestibulum) of wild Afri- 
can elephants which did not contain viral 
inclusion bodies histologically This her- 
pesvirus is now considered to be a novel, 
indigenous herpesvirus of African ele- 
phants, possibly latent in lymphocytes of 
the vaginal lymphoid follicle-containing 
patches. These lymphoid patches periodi- 
cally become hyperplastic and occasionally 
ulcerate (Munson et al., 1995) which then 
may afford transmission of the virus to an- 
other host. The additional findings in our 
report here indicate that there is a second 
related herpesvirus identified in pulmo- 
nary nodules obtained from healthy wild 
African elephants that has 100% protein 
identity (in the DNA polymerase gene re- 
gion) with the previously reported virus 
that was lethal for the two African ele- 
phants with endotheliotropic disease. Al- 
though currently limited by data obtained 
from a small portion of the viral genomes, 
and given the obvious need for additional 
viral sequence data, the results obtained to 
date may indicate that African elephants 
can harbor two novel herpesviruses, one 
that can cause fatal endotheliotropic dis- 
ease in Asian elephants and the other in 
African elephants. Currently, intensive 
work is underway to further elucidate the 
epidemiology of these two herpesviruses 
which   are   either  carried  latently,   or  as 
RICHMAN ET AL.?A NEW DISEASE OF ELEPHANTS CAUSED BY ENDOTHELIOTROPIC HERPESVIRUSES 
chronic low-level infections in African el- 
ephants. 
In addition to the Asian elephant from 
Switzerland (Ossent et al., 1990), the virus 
of which we determined is homologous (in 
the terminase gene region) with our other 
nine Asian elephant cases from North 
America, we have confirmed by PCR and/ 
or morphological techniques similar her- 
pesvirus cases in five Asian elephants from 
Germany, Netherlands, and Israel (Rich- 
man et al., 1999b). The implications of the 
herpesvirus in wild African elephants are 
unknown but no herpesvirus-associated ill- 
nesses have been observed or reported 
through the diagnostic laboratory in On- 
derstepoort. Republic of South Africa. The 
existence of any elephant herpesviruses in 
Asian elephants in the wild is currently un- 
known. 
Although our findings to date implicate 
the African elephant as a potential reser- 
voir of herpesviruses that can cause dis- 
ease in the two elephant species, direct 
proof of transmission has not been estab- 
lished between contact and exposed ani- 
mals. Considerably more epidemiological 
work is required to clarify the potential 
carrier status of the Asian elephant for ei- 
ther of these viruses. Some of these factors 
will depend on cultivating the viruses, or 
by the development of serological tests so 
that elephants previously exposed to these 
viruses can be identified. It might be im- 
portant to minimize the co-mingling or so- 
cializing of Asian and African elephants at 
the same facility or between facilities as 
this may increase the risk of the transmis- 
sion of herpesviruses particularly in breed- 
ing operations, or where new elephants 
are being introduced. 
ACKNOWLEDGMENTS 
Funded by NIH grant No. 1 K08 AI01526- 
01, the Smithsonian Scholarly Studies Program, 
the Kumari Elephant Conservation Fund, and 
Friends of the National Zoo. The authors thank 
R. Garber, D. Nichols, V. Bonshock, D. Fi- 
scher, A. Bratthauer, N. Spangler, L. Cheng, K. 
Clark, J. Sutton, E. Lamirande, J. Lehnhardt, 
N. Pratt, M. Bush, J. Block, the elephant keep- 
er staff at the National Zoological Park, P. An- 
gle, T- Cohen, S. Feldman, S. Mikota, R. Mir- 
kovic, J. d'Offay, R. Eberle, G. Letchworth, K. 
Steele, B. Connolly, P. J?hrling, J. C. Zong, M. 
Lucskay, J. Cannon, J. Yang, R. Ambinder, J. 
Nicholas, A. Hess, L. Poole, D. Ciufo, W. Gib- 
son, A. Ruebel, P. Ossent, F Osorio, S. Kania, 
M. Kennedy, E. Dierenfeld, T. McNamara, J. 
Trupkiewicz, L. Munson, L. Lowenstine, B. 
Rideout, I. Stalis, J. St. Leger, J. Fickel, J. Par- 
rott, K. Emanualson, D. Taylor, J. Jenkins, R. 
V. Ferris, D. Scott, G. Coleman, B. Schmitt, A. 
D. Alstad, J. Gaskin, D. Olson, M. Keele, A. 
Schanberger, E. Jacobson, L. Gage, Marine 
World Africa-USA, L. Bingaman-Lackey, N. 
Kriek, A. Greenwood, W. Lindsay, G. West, R. 
Isaza, L. Reeve Peddie, J. Peddie, J. Williams, 
T Williams, elephant keeper staff at Ringling 
Brothers, San Diego Zoo and Wild Animal 
Park, Center for Reproduction of Endangered 
Species (GRES), New York Wildlife Conser- 
vation Society, Lincoln Park Zoo, Dickerson 
Park Zoo, African Lion Safari, Tulsa Zoological 
Park, Fort Worth Zoo, Indianapolis Zoo, Dallas 
Zoo, Houston Zoo, Oakland Zoo, Have Trunk 
Will Travel, National Veterinary Services Lab- 
oratories, and California Veterinary Diagnostic 
Laboratory. 
LITERATURE CITED 
BOIKE, S. C, M. PUE, P. R. AUDET, M. I. FREED, A. 
FAIRLESS, B. E. ILSON, AND N. J. D. K. ZARIFFA. 
1994. Pharmacokinetics of famciclovir in subjects 
with chronic hepatic disease, journal of Clinical 
Pharmacology 34: 1199-1207. 
FOWLES, S. E., AND D. M. PIERCE. 1989. High per- 
formance liquid Chromatographie method for the 
determination of 9-(4-hydroxy-3-hydroxymethyl- 
but-l-yl) guanine (BRL-39123) in human plasma 
and urine. Analyst 114: 1373-1375. 
JACOBSON, E. R., J. P. SUNDBERG, J. M. GASKIN, G. 
V. KOLLIAS, AND M. K. O'BANION. 1986. Cuta- 
neous papillomas associated with a herpes\4rus- 
like infection in a herd of captive African ele- 
phants. Journal of the American Veterinary Med- 
ical Association 189: 1075-1078. 
McCuLLY, R. M., P. A. BASSON, J. G. PIENAAR, B. J. 
ERASMUS, AND E. YOUNG. 1971. Herpes nodules 
in the lung of the African elephant (Loxochmta 
africana). Onderstepoort Journal of Veterinary 
Research 38: 225-236. 
METZLER, A. E., P. OSSENT, F. GUSGETTI, A. RUBEL, 
AND E. M. LANG. 1990. Serological evidence of 
herpesvirus infection in captive Asian elephants 
(Elephas niaximus). Journal of Wildlife Diseases 
26: 41-49. 
MiKOTA, S. K., E. L. SARGENT, AND G. S. RANELAK. 
1994. Medical management of the elephant. In- 
12 JOURNAL OF WILDLIFE DISEASES, VOL. 36, NO. 1, JANUARY 2000 
dira Publishing House, West Bloomfield, Michi- 
gan, pp. 54?58. 
MOORE, P. S., AND Y. CHANG. 1995. Detection of 
herpes\4rus-hke DNA sequences in Kaposi's sar- 
coma in patients with and without HIV infection. 
New England Journal of Medicine 332: 1181- 
1185. 
MuNSON, L., W. KARESH, S. SHIN, J. BALKE, P. CAL- 
LE, R. CAMBRE, M. CRANFIELD, S. CITINO, AND 
R. JUNGE.  1995. Lymphoid foUicular VTilvitis in 
African {hoxodonta africana) and Asian (Elephas 
maximus) elephants. Journal of Zoo and Wildlife 
Medicine 26: 353-358. 
OSSENT,   P.,   F.   GUSCETTI,   A.   E.   METZLER,   E.   M. 
LANG, A. RUBEL, AND B. H?USER. 1990. Acute 
and fatal herpesvirus infection in a young Asian 
elephant (Elephas maximus). Veterinary Pathol- 
ogy 27: 131-133. 
RICHMAN, L. K., R. J.  MONTALI,  R.  L.  GARBEB,  M. 
A. KENNEDY, J. LEHNHABDT, T. HILDEBRANDT, 
D. SCHMITT, D. HARDY, D. J. ALCENDOR, AND 
G. S. HAY'WARD. 1999a. Novel endotheUotropic 
herpesviruses fatal for Asian and African ele- 
phants. Science 283: 1171-1176. 
 , , T. HILDEBRANDT, J. FICKEL, D. 
SCHMITT, AND G. S. HAY-WARD. 1999b. Status of 
a ne\v, fatal herpesvirus disease in elephants in 
North America and Europe. Verhandlangsber- 
icht des 39 International Symposium ?ben Er- 
krankungen der Zoo und Wildtiere, Wien 39: 17? 
21. 
SCHMITT, D. L., AND D. A. HARDY. 1998. Use of 
famciclovir for the treatment of herpes\4rus in an 
Asian elephant. Journal of the Elephant Manag- 
ers Association 9: 103?104. 
SCHULLER, W, S. CERNY-REITERER, AND R. SILBER. 
1990. Evidence that the sheep associated form 
of malignant catarrhal fever is caused by a her- 
pes\4rus. Zentralbl Veterinarmed 37: 442?447. 
STELLE, K. E., F. Y. SCHULMAN, H. MENA, AND E. 
O. STRIMPLE. 1997. Rhabdoid tumor in the brain 
of a dog. Veterinary Pathology 34: 359?363. 
VERE HODGE, R. A., D. SUTTON, M. R. BOYD, M. R. 
HARNDEN, AND R. L. JARVEST. 1989. Selection 
of an oral prodrug (BRL 42810; famciclovir) for 
the antiherpeswrus agent BRL 39123 [9-(4-hy- 
droxy-3-hydroxymethylbut-l-yl) guanine; penci- 
clo\4r]. Antimicrobial Agents and Chemotherapy 
33: 1765-1773. 
Received for publication 19 February 1999.