Vet Pathol 311: 555-559 ( 1993)

Clinical and Histologic Characterization of Cutaneous

Reactions to Stings of the Imported Fire Ant
(Solenopsis invicta) in Dogs
P. M . RAKICH , K. S . LATIMER, M. E. MISPAGEL,
AND W . L. STEFFEN S

Athens Diagnostic Labora tor y (PMR), Departmen t of Veteri nary Path ology (KSL),
Enviro nmental Toxicology Program (MEM), and Departm ent of Anatomy and Radi ology (WLS),
College of Veterinary Medicine, University of Geo rgia, Athens, GA

Abstract. Four adult dogs received experimentally controlled stings in the dor solateral abd ominal skin by
imp ort ed fi re ants (Solcnopsis invicta i . Th e sites were exa mined grossly 15 minut es and at I , 2, 3, 4, 5, 6, 24,
48, and 72 hour s and histologically 15 minutes and 6, 24, 48, and 72 hour s after stinging. The initial gross
lesions at 15 minutes were swelling and erythema, and the microscopic changes were vascular congestion and
superficial dermal ede ma. By 6 hou rs, the lesions consiste d of bright erythema tous pru ritic papul es characterized
microscopically by a band of full thickness derm al necrosis and inflammat ion . By 24 hou rs and continuing to
the end of the study at 72 hour s, the sites appeare d completely norm al grossly. Biopsies taken 24, 48, and 72
hour s after stings conta ined microscopic changes similar to those present at 6 hours after stings. Th ese hist ologic
changes are unlike those described for hum an beings stung by import ed fire ants. In hum an beings, fire ant
stings are characterized histologically by an initial superficial vesicle that evo lves into a sterile pustule.

Key words: Dogs; fi re ants; fire ant stings; skin; So lenopsis invicta.

The imported fire ant iS olcnopsis invicta) wa s intro-
duced from South America into the United States by
ship sometime in the 1930s o r 194 0s through the port
o f Mobile, Alabam a." Since th en , th e ant' s range has
spread to in vol ve a p p rox imate ly 230 million ac res in
ten southern sta tes - A la ba m a , Fl o rida , G eorgia , Lou-
isiana, Mi ssissippi , North a nd South Caro lina, T exas,
T ennessee , and Arkansas.' Fire ants inhabit urban as
well as rural settings and cause damage to human be-
ings, vario us anim al s, cro ps, farm machinery, and elec-
tri cal in sul ation. i-! " When a nest is di sturbed, hundred s
o f a n ts swa rm out, cover a ny nearby o bjects , and d e-
liver num erous stings. Their name is d eri ved from th e
in itia l painful burning se ns ation elic ite d by th e stings.'
When stinging, fire ants attach them selves by pinching
the sk in with th eir mandibles , arch their body, and
th en inj ect 0.04-0.11 J,Ll of venom through an abdom-
inal stinge r (Fig. I). T he ants th en typ icall y pi v ot a bo ut
th eir heads and d el iver multiple ad d itio na l stings. >"
Thirty percent to 60% o f human beings in urban a reas
infested by imported fire a n ts are st u ng e ve ry yea r.8.9
Various local and systemi c reaction s can d evelop in
human beings st u ng by imported fire ants, and these
ha ve been we ll cha rac te rize d cl inicall y a nd micro scop-
icall y.5.8,9,' 5
The combinati on o f th e indiscriminately inq u isitive
nature o f d ogs and th e aggressi veness of th e imported
555
fire ant suggests th at dogs living in infested areas will
be stung. Because skin biopsies are frequentl y sub-
mitted by vete rina ria ns as part of a diagnosti c workup
for d ermatopathies of uncertain ca use , th e hi st ologic
cha nges assoc iate d wit h imported fire ant stings must
be kn own to co rrec tly include o r ex cl ude th ese sting
reactions in a differential histopath ologic d iagnosis,
However , there is no information in th e vet erinary
litera t ure concerning th e gross appearance or hi stologi c
c ha nges in the skin o f d ogs resulting from stings of fire
a n ts. The purp o se o f thi s study was to in vesti gate th e
gr o ss a ppe a ra nce a nd mi croscopic cha nges in canine
skin a t va rio us time period s foll owing fire a n t enven -
o m a tio n .
Materials and Methods
Dogs
Four privately owned (by P. M. Rakich and K. S. Latimer)
dogs of mixed breed ing and sex were used for the study: dog
No. I, 8.5-year-o ld castrated male terrier mix; dog No. 2,
> 10-year-old male American Foxhound ; dog. No. 3, > 10-
year-old male Border Collie mix; dog No.4, 2-year-old spayed
fem ale Labrador Ret riever. All dogs were clinically health y
and free of external skin lesions as determi ned by physical
exam ination. Th e dogs were fed a commercia l diet and pro-
vided water ad libitum . T he hai r was clipped from the dor -
solatera l abdo minal area using a #40 blade.
556 Raki ch et a l. Vet Path ol 30:6. 1993

Fig , I. Scanning electro n mi crograph . Skin; dog . A fire Fig, 2. Skin; dog No. 3, 15 minutes after imported fire
a nt is in the process ofstinging. Not e mandibular att achment, a nt sting. Super ficial dermal co llage n fibers a re wid ely spaced
a rched bod y, a nd stinger pen etrati on of th e epide rm is. Bar = d ue to derm al ede ma . Vessels a re co nges ted . HE . Ba r = 75
.25 111m . /l m .

Fire ants iSolenopsis invictas and stings after the sting and consi sted of erythe ma and swellin g
measuring less than I em in diameter. Central blanch-
Wild -cau ght fire ants were used for th e st udy . Fire a nt
mounds were di sturbed , ind ucing the ants to swa rm. Both
ing was observed in so me of the reacti ons. Eryth ema
a nts a nd red clay were shoveled int o glass jar s that were lined persisted and swelling continued to enlarge to a max-
wit h talc to prevent the ants from esca ping. Indi vidual a nts im um diameter of 1.5-2 em at approx imately 6 hours
were selected for use a t random and handled gently with after the stings. By 24 hours, erythema was minimal
lat ex gloves a nd forcep s. Each fire ant was all owed to sting to barely discernible and was the only change evident
o ne cutaneo us site; eac h dog recei ved a sting at five site s of in two dogs (Nos . I, 2). On e dog (No .3) had an ap -
th e shaved area and the sting sites were enc ircled by ind elibl e par ent pustule (approximatel y 2 mm in diameter) and
marking pen . The do gs were o bse rve d for 15 minutes after another dog (No.4) had superficial wrinkling of the
th e stings and then hourl y for 6 hours to record th e gro ss skin , suggestive of a collapsed vesicle or pust ule. By
a ppea ra nce of th e reactions and to look for evidenc e of sys-
48 hours, no erythema was evident. Two dogs (Nos .
temic reacti ons.
2, 4) had small palpable dermal ind urations, but no
Skin biopsies abnormalities were gros sly or palpably evident in the
other two dogs . Lesions seemed to be pruritic or irri-
Routine skin biopsies wer e tak en usin g local subcuta neo us
tating, because two of th e dogs (Nos . I, 4) scratched
a nalgesia a nd a 6-m m biopsy pun ch. Biop sies were obtained
from eac h dog prior to th e stings a nd a t 15 minutes a nd 6,
and chewed the sites . Had the stings not been marked,
24 ,48 , a nd 72 hours after fire a nt st ings. Biop sies were fixed it would hav e been impossible to visually locate the
in neu tral buffer ed 10% formal in and processed routinely, sting sites 48 hours aft er the stings.
a nd 3-/l m sec tions were cut a nd stai ned by th e hematoxylin Skin biopsies taken prior to fire ant stings were de-
a nd eosi n (HE) a nd Gie msa techniques. Giems a stai ning fa- vo id of inflammation in all four dogs. Th e dermis con-
cilita ted differentiating eos ino phils from neut rophils. tain ed occasional mast cells, most of which were scat-
tered around superficial dermal blood vessels. Mast
Resul ts cell numbers averaged three cells per high-power (45 x
None of the dogs exhibited ev ide nce of pain (i.e., no objective) field of view.
flinching or vocalization) in response to the fire ant Fifte en minutes aft er stings, the most consistent le-
stings. In contrast, th ey frequ entl y responded when sion was superficial dermal va scular congestion (4/4
lidocaine was injected subcutaneousl y for local anal - dogs). Granulocyte pa vementing was evident in two
gesia . dogs (Nos. 3,4), and emigration of leukocytes (mostly
Gross changes developed within 5 to 10 minutes eosinophils) was seen in one of these two dogs (No.4).
Vel Path o l 30:6. 199 3 Ca n ine React ion s to Fire An t St ings
Superficia l derm al co llage n fibers were widely sepa-
ra ted , and ede ma fluid was so me times visible as pale
eosinophilic lakes or as basophilic gra nular materi al
(4/4 dogs) (Fig . 2). A few sma ll ar eas o f hem orrhage
were see n in two biopsy specimens (dog No s. 2, 3).
At 6 hours aft er stinging, th e epide rm is was un-
cha nged in d og Nos. 2 and 3 and m inimall y thi cken ed
in dog Nos. I a nd 4. Intracellul a r ede ma was m ild in
d og NO. 4. Dermal lesion s in all four dogs were a lin ear
band of full thickness dermal necrosi s in the ce nter of
th e biopsy spec ime n (Fi g. 3). The cente r of th e lesion
co nsisted o f a n irregular area of coagulative necrosis
of the dermis, includ ing ad nexa l structures , a nd oc-
cas iona lly o f th e supe rficial subcutis. Int en se in fla m -
mati on , co m posed o f swirling nucl ear debris a nd pyk-
noti c degen erati ve in fl a m matory cells, sepa rated dermal
co llagen fibers and form ed a border surro undi ng th e
necroti c cente r (Fig. 4). Giemsa-stai ned sec tio ns in-
di cat ed th at eos ino phils co m prised man y to most o f
th e gra n ulocytes.
Dermal tissue surro und ing necr oti c foci was ed em-
a to us a nd infiltrat ed diffu sely or in a per ivascul ar pat -
tern with eos ino phils primaril y ad m ixed with fewer
neutrophil s, lym ph ocytes, a nd macrophages. Vessels
were packed with gra n ulocy tes or had pr om inent gran-
ulo cyte pavem enting. Sma ll ar eas of hem orrhage wer e
scatte red through out th e dermis in one dog (No. 2).
Mas t cells with very few gra n ules were ide nti fied in
two dogs (Nos. I, 2).
At 24 hours after st ings, the epide rm is was m ildl y
thi cken ed . In d og No .3, th e epidermi s was diffus ely
and uniforml y necroti c a nd sepa ra ted from th e dermis.
Derm al cha nges were si mi lar to th ose seen a t 6 hours
a fter st ingi ng. The maj orit y of th e recogni zable inflam-
mat ory cells were eos ino phils.
At 48 hours aft er stinging, th e epide rm is was mildly
hyperplasti c. In dog No. 2, the epide rm is was diffusely
necroti c a nd th e dermis was covered on ly by stratum
co rne um . Nec rosis o f th e derm is in th e center of th e
biop sy speci me n exte nde d do wn to th e subc uta neous
fat. Man y o f th e inflammatory cells had small pykn oti c
nucl ei and diffu sely eos ino philic cyto plas m and co uld
not be ide nti fied . Sca tte red cells had a few eosino ph ilic
granul es, ind icating th at they were necr oti c eosino -
phi Is.
At 72 hours after stinging, th e epide rm is was mildl y
hyperplastic in all four dogs. Th e epide rm is of one d og
(No . 3) also had m ild interc ellular ede ma with exo-
cytos is o f eos ino phi ls a nd an int raepidermal pu stul e
co nta in ing erythrocy tes , pykn ot ic granulocytes, and
occasional eo sino phils. Changes in th e dermis wer e
similar to those see n at 24 and 48 hours with a band
o f full-th ickn ess necrosis exte nd ing down to a nd, in
two dogs (Nos. I, 4), in vol vin g th e subc utis. The ma-
j ority of necr ot ic tissue was in filtra ted with large num-
557
bers o f inflammat ory cells tha t wer e also necroti c a nd
co uld not be identified . The surro unding via ble derm is
was co nges ted a nd ede ma to us a nd co ntai ne d mi xed
inflamm at ory cells (lymphocytes, plasm a cells, m ac-
ro phages, eo sino ph ils, and neutrophils) loosely d is-
tributed a mo ng co llagen fiber s or a ro und blood vesse ls.
Discussion
The initi al res po nse of dog s stung by im po rted fire
ants cons isted of erythe ma and swelling, whi ch devel-
oped within 10 to 20 minutes a fter stinging a nd re-
so lved wit hin I hou r. Micr oscopi cally, cha nges at 15
minutes a fter sti ngi ng co nsis ted of co nges tio n, derm al
ede ma, a nd mild gra nulocy tic pa vem enting in super-
ficial derm al blood vessels. G ro ssly and histologicall y,
th e lesion s were sim ilar to th e initial react ion s in hu-
man bein gs stung by imported fire ants a nd were con -
siste nt with whea l a nd flare react ion s.v"
A fter resolution of th e initia l whea l and fl ar e reac-
tion , th e sting sites became ery the ma to us a nd eleva ted
to form bright red papules by 4 to 6 hours after stinging.
So me of th e d ogs scra tc hed or licked th ese lesion s,
suggesti ng th at th ey were pruritic or ir rita ti ng. By 24
hours, erythe ma was mild , a nd altho ugh papules were
no longer present, indurati on was palpable. By 48 hours
after stinging, sites could be identified only because o f
th e inde lible ink marks. If th e stinging sites had not
been ma rked , th ey co uld not ha ve been biop sied with
certa inty. T hese gross cha nges a re unlik e th ose in hu-
man bein gs st ung by imported fire an ts. A fte r resolu-
tion o f th e immediate wh eal a nd flar e rea ct ion withi n
a n hour, th e lesion in human bein gs evolves int o a
clear fluid-filled ves icle a nd th en into a sterile pustul e
a t th e site o f th e sting by 24 ho urs.>? T he pu stul e re-
main s for 3 to 10 da ys th en ruptures as th e epide rm is
sloughs o ff; th e lesion th en reepithelializes.>" The le-
sions can be int en sely pruritic, and seco nda ry bacteri al
in fection is a co m mo n co m plica tion in hum an bein gs
whe n fire a nt sti ngs become excoriated ." Less freque nt-
ly, per son s stung by im po rted fire ants de velop la rge
local rea cti on s cha racterized by ery t he ma, ede ma, in-
durati on , pain, and pruritus a t th e site o f th e sting.s-?
Vesicles a nd pu stul es were not prominent in an y o f
th e d ogs in thi s study. T he epiderm is of dogs is thin ner
th an th at o f human s, 17 whic h co ntributes to th e ephe m-
era l nature of pu stul es or ves icles in d ogs. However,
at 24 hours, one sting site resembled a co lla pse d pu stul e
or ves icle and a seco nd a ppea red to be a tin y superficia l
pu stule. No lesion co m pa ra ble to a la rge local reac tion
as obse rve d in human bein gs, develop ed in any of th e
dogs.
Stings of imported fire ants in th e ex perime ntal dog s
pr oduced a disti ncti ve necr oti zin g derm at iti s at 6, 24 ,
48, and 72 hours a fter sti ngi ng. The m icro scop ic le-
sions were frequentl y linear, in volvin g th e full thi ck-
558 Ra kich et al. Vet Pat hoi 30:6. 1993

Fig. 3. Skin; dog No.4, 6 hours a fter im port ed fire ant sti ng. A linear band o f necro sis (arrows) extends from epiderm is
to subcuta neo us fat. H E. Bar = 375 li m .
Fig . 4. Skin; dog No .4, 6 hours a fter import ed fire ant sti ng. Degenerative eos ino phils and neut roph ils border the
necro tic tissue. Gicmsa . Bar = 15 li m.

ness o f th e dermis and exte nd ing va rio us depths into
th e subcutis. Infl ammati on wa s inten se and primaril y
granulocyti c, but th e majority of th e cells were to o
degen erate to be identified as eosinophils or neutro-
phils. Both ce ll typ es wer e evide nt in the adj acent vi-
a ble dermis. Had sting sites been biopsied soo ne r a fte r
th e sting, necr osi s wo uld ha ve been less severe a nd
d iffuse, a nd inflammato ry ce lls mi ght ha ve been m o re
identi fiabl e. Ho wever, a 6- ho ur post- st ing inter val was
chosen becau se th e rea cti on sites we re clin icall y o b-
vio us, making th em a lik ely bi opsy site . In additi on ,
this tim e wa s chosen to account for th e time that might
ela pse between the point when an owner would notice
a lesion (by 4 hours after stinging, well developed red
papules wer e visible) a nd wh en th e dog co uld be ex-
a mi ned and a skin bi opsy taken by a ve te rina ria n (2
hours lat er). Although necrosis is described as th e re-
spo nse to va rio us art hro po d sti ngs and bites in d ogs,
th e hi stologic cha nges are poorl y cha racterized a nd in-
co m plete ly described . P" The cha racteristic hist ologic
cha nge in human beings st ung by fire a nts co nsists o f
a prominent su pe rficia l pu stu le pr ojecting above th e
skin surface . It is bordered above by a thin ro of o f
stratum co rne um with a layer of a few epide rm a l cells.
The floor o f th e pu stule co ns ists o f epide rm is except
at th e ce nte r, wh ere th e epiderm is is rep laced by ne-
cro tic dermal ti ssu e, ben eath wh ich th e supe rficia l der-
m is is d iffusely infiltrat ed by m ixed inflammat o ry
cells.>"
The veno m of th e imported fire a nt d iffer s in its
chem ica l co m positio n from tha t of other hym en op -
teran sti nging insects. U n like th e ve no ms ofbees, was ps,
and hornet s, which a re primaril y aq ueo us so lutio ns
co nta in ing prot ein s," th e ve no m o f th e imported fire
ant is wat er in so luble" a nd co m posed primari ly of a
un ique a lka loid (sol enopsin Ap ·1 2 with o nly a small
aqueous fraction co nta in ing so luble prot eins.v' ? The
a lka loid portion of th e ve no m , whic h co ns titutes 95%
(by weig ht) o f th e ve no m , is cyto toxic, hem ol yt ic, fun-
gicidal, insecti cidal, a nd bact eri cidal .t-v' -? The pro -
nounced dermal necr osis developing at th e site of th e
sti ng is th ought to be directl y a ttri buta ble to th e a l-
kal o id s. ' T opi cal tr eatment of sting sites with an tib i-
ot ics, co rticos teroi ds, o r antihi stamines a nd treatm ent
Vet Path ol 30 :6. 199 3 Canine Reactions to Fire Ant Stings
with oral antihistamine failed to alter th e clinical ap-
pearance or developm ent of lesions in human vo lun-
tcers.' Similarl y, tr eatm ent of stings with topical co r-
ti co st eroid s o r a n ti bac te ria l so lutio n, parenteral
corticosteroi ds , anti hista mi ne , or epi ne phrine a nd pre-
tr eatment with top ical o r pa renteral co rticoste ro ids did
not alte r th e clin ical co urse or pathol ogic resp on se to
im ported fire ant sti ngs in rabbits. 14 These observa tio ns
su ppo rt th e th eory that the in iti al necrotizing response
to im po rted fire ant sti ngs is due to a d irect effect of
th e alka loids rather th an to immunologicall y m edi ated
events .
In addi tio n to th e local cuta neous rea cti on s, life-
threatening systemic a llergic reacti on s ca n occur, de-
velo ping up to several hours a fte r th e sti ng, in an es-
timat ed 0.6- 1.0% o f people stung by imported fire
ants."-'> These reacti on s are m edi at ed by se ru m IgE
d irect ed agai nst prot ein all ergen s in th e ve no m . Al-
th ou gh prot ein s account for o nly about O. I% by we ight
of im ported fire ant ve no m , th ese proteins rather th an
th e alkalo id s a re res po ns ible for inducti on o f ana phy -
lacti c rea cti on s from sti ngs of im po rted fire ants.' The
ve no m co ntai ns at lea st three proteins, two o f which
are also found in o ther hym en opteran ve no ms, i.e. , a
ph osphol ipase and hyaluron idase.v! ' No evide nce of
a na phy laxis was see n in th e d ogs st ung by im po rte d
fire a nts in th is st udy , but a llerg ic reactio ns ca n occ ur
a nd ha ve been d ocumented in d ogs stung by othe r
hym en opteran insec ts. "
Acknowledgements
This research projec t was funded by a grant (29-26 -G R207 -
006, 9 1-RA K) from the University of Georgia Veteri nary
Medical Experimen t Station Fire Ant Program. We thank
M. Kathy Davis, HT(ASC P), and Mary B. Ard, HT (ASCP ),
for technical suppo rt.
References
Adro uny GA , Derbes VJ , Jung RC: Isolat ion of a he-
mo lyt ic co m ponent of fire ant venom . Science 130:449,
1959
2 Baer H, Liu TY , Ande rso n MC, Blum M, Schmid WH ,
Jam es FJ: Prot ein com ponents of fire an t venom (Sole-
nopsis in victa ). Tox ico n 17:397-405 , 1979
Requ est reprints from Dr. P. M. Rak ich , Athens Diagnostic Labo ratory, Co llege of Veterinary Med icine, Un iversit y of
Georgia, Athens, GA 30602-7383 (USA).
559
3 Blum MS: Che mis try and properties o f fire an t veno ms .
Proc Symp Import ed Fire Ant, June 7- 10, Atlanta, GA ,
pp. 116-11 7, 1982
4 Blum MS, Walker JR , Callahan PS: Chem ical, insec ti-
cidal, and antibiotic pro perties offire ant veno m. Scienc e
128:306-307, 1958
5 Ca ro MR, Derbes VJ, Ju ng R: Skin respo nses to the
sti ng of the im ported fire an t iSolenopsis sacvissim ai.
Arch Dermatol 75: 475-488, 1957
6 Cowe ll AK , Co well R L, Ty ler RD , Nieves MA: Seve re
syste mic reac tions to Hymen optera sti ngs in three dogs.
J A m Vet Med Assoc 198:1014-1 01 6,1 991
7 Da vids on NA , Ston e N D: Im ported fire an ts. In: Erad-
ication o f Exoti c Pests, cd . Da hlsten DL and Garcia R,
pp. 196-2 17. Ya le University Press, New Haven, CT,
1989
8 deShazo RD , Butcher BT, Ban ks WA: Reacti on s to the
stings of the imported fire ant. N Engl J Med 323 :4 62-
466 , 1990
9 deShazo R D, Griffing C, Kwan T H, Banks WA , Dvorak
H F: Dermal hypersensi tivit y rea ction s to imported fire
ants. J Allergy Clin Immunol 74:841-847, 1984
10 Green HB: Biology and contro l of the im ported fire an t
in Mississippi . J Econ Ento rno l 45:593-597 , 1952
II Hoffm an DR : Allergens in Hymen optera veno m . XV II.
Allergic co mpo nents of So lcnopsis invicta (imported fire
an t) veno m. J Allergy Clin Immunol 80: 300-306, 1987
12 Mac Connell JG , Blum MS: Alkaloid from fire ant ven-
om : identification and synt hesis . Science 168:840- 84 1,
1970
13 Mull er G H, Kirk R W, Sco tt DW: Cutaneous parasitol-
ogy. In: Sma ll Ani ma l Dermatology, 4th ed ., pp . 347-
426. WB Saunders, Ph ilad elphia, PA, 1989
14 Par rino J , Kandawalla NM , Lockey R F: Treatmen t of
the local skin response to imported fire ant stin g [ab-
stract ]. J Allergy Clin Im mu nol 63: 135, 1979
15 Rhoades RB, Schafer WL , Sch mi d W H, Wubbena PF,
Dozier RM , Townes AW, Witt ig HJ : Hypersensitivity
to the imported fire ant. J Allergy C1i n Irnm uno l 56: 84-
93 , 1975
16 Tsc hinkel WR : H istory and bio logy of fire ants. Proc
Symp Imported Fire An t, June 7- 10, Atlanta , GA , pp .
16-35, 1982
17 Yager JA , Scott DW : T he ski n and appendages. In: Pa-
thology of Domestic Anima ls, ed . J ubb KVF, Kennedy
PC, and Palm er N, 3rd ed ., vo l. I, pp. 49 1-502. Aca -
de mic Press, O rla ndo , F L, 1985