Acta Tropica 93 (2005) 85–95

Neutralization of four PeruvianBothropssp. snake venoms bypolyvalent antivenoms produced in Peru and Costa Rica:

preclinical assessment

Ermila Rojas, Lil Quesada, Viviana Arce, Bruno Lomonte,Gustavo Rojas, Jose Marıa Gutierrez∗

Instituto Clodomiro Picado, Facultad de Microbiolog´ıa, Universidad de Costa Rica, San Jos´e, Costa Rica

Received 2 November 2003; received in revised form 1 September 2004; accepted 10 September 2004Available online 11 November 2004

Abstract

Envenomations after bites inflicted by snakes of the genusBothropsconstitute a public health hazard in Peru, and theintravenous administration of equine-derived antivenoms represents the only scientifically validated treatment. This study presentsa preclinical assessment of the efficacy of two whole IgG antivenoms, prepared in Peru and Costa Rica, to neutralize the mostrelevant toxic effects induced by the venoms ofBothrops atrox,B. brazili,B. barnettiandB. pictusfrom Peru. Peruvian antivenomis produced by immunizing horses withBothropssp. venoms from this country, whereas the production of Costa Rican antivenomi -forming,m om priort values ofe lization ofl ofb enom.I a,a©

K

f

t in-lap-a

0d

nvolves immunization with venoms from Central American snakes. The neutralization of lethal, hemorrhagic, edemayotoxic, coagulant and defibrinating activities was evaluated in assays involving incubation of venom and antiven

o testing. Both antivenoms were effective in the neutralization of these effects, with quantitative variations in theffective dose 50% depending on the effects being studied. Peruvian antivenom was more effective in the neutra

ethality induced byB. atroxandB. barnettivenoms. However, Peruvian antivenom failed to neutralize coagulant activityB.arnettivenom and edema-forming activity ofB. brazilivenom, whereas neutralization was achieved by Costa Rican antiv

t is concluded that an extensive immunological cross-reactivity exists betweenBothropssp. venoms from Peru and Costa Ricnd that both antivenoms are effective in the neutralization of these four venoms in a rodent model of envenoming.2004 Elsevier B.V. All rights reserved.

eywords:Snake venom; Antivenom;Bothropssp. venoms; Neutralization; Peru

∗ Corresponding author. Tel.: +1 506 229 0344;ax: +1 506 292 0485.

E-mail address:jgutierr@icp.ucr.ac.cr (J.M. Gutierrez).

1. Introduction

Peru has a rich and diverse herpetofauna thacludes venomous snake species of the families Eidae (16 species ofMicrurus and the pelagic se

001-706X/$ – see front matter © 2004 Elsevier B.V. All rights reserved.oi:10.1016/j.actatropica.2004.09.008

86 E. Rojas et al. / Acta Tropica 93 (2005) 85–95

snakePelamis platurus) and Viperidae (15 species)(Campbell and Lamar, 1989). Snakebite envenoma-tions represent a public health problem in this coun-try (Zavaleta and Salas, 1996). The vast majority ofsnakebites in Peru are inflicted by species of thegenusBothrops(family Viperidae) (Zavaleta and Salas,1996). B. atrox, B. brazili, andB. bilineatusare dis-tributed in the tropical rainforests located in the easternpart of the country, whereasB. barnettiandB. pictusare found in the western dry coastal regions (Campbelland Lamar, 1989; Fan and Cardoso, 1995). In addi-tion to these viperid species, which represent the mainproblem, a number of envenomations are inflicted byBothrocophias(formerly Porthidium)hyoprora, Both-rocophias microphthalmus, Lachesis muta, from thefamily Viperidae, and various species ofMicrurus,from the family Elapidae (Campbell and Lamar,1989). The rattlesnakeCrotalus durissus terrificus(family Viperidae) is distributed only in a lim-ited region in the southeastern part of the country(Campbell and Lamar, 1989), and causes very fewaccidents.

The Centro Nacional de Produccion de Biologicosof the Instituto Nacional de Salud, in Lima, manufac-tures an ‘antibothropic’ polyvalent antivenom, used totreat envenomations byBothropssp. (Theakston andWarrell, 1991; Incio-Ruiz et al., 1993; Zavaleta andSalas, 1996). It is a whole IgG preparation producedfrom the plasma of horses immunized with the ven-oms ofB. atrox, B. pictus, B. barnetti, B. brazili andB d,L thevP -t alsou an-t inC -p an-t era y oft beenp Ri-c va-l em en-of

2. Materials and methods

2.1. Venoms and antivenoms

Venoms from the following species were used:Both-rops atrox,B. brazili,B. barnettiandB. pictus. Venomswere obtained from specimens collected in various lo-cations in Peru and brought to the serpentarium of theInstituto Nacional de Salud, Lima. Venom samples arepools from many individuals of both sexes and differ-ent ages, with predominance of adult specimens. Af-ter extraction, venoms were immediately lyophilized.In some experiments, a venom pool ofB. asperfromadult specimens collected in Costa Rica was used.The following antivenoms were tested: (a) polyva-lent ‘bothropic’ antivenom (liquid, batch 009030; totalprotein concentration: 6.7 g/dL), produced by CentroNacional de Produccion de Biologicos, Instituto Na-cional de Salud, Lima, Peru. It is obtained by immu-nization of horses with a mixture of the venoms ofB. atrox, B. pictus, B. barnetti, B. brazili andB. hy-oprora (Instituto Nacional de Salud, Lima, personalcommunication) followed by ammonium sulfate frac-tionation of horse plasma to obtain a whole IgG prepa-ration (Organizacion Panamericana de la Salud, 1977).This antivenom is standardized as to have a potencyof 2.5 mgB. atrox venom neutralized per mL of an-tivenom. (b) Polyvalent (Crotalinae) antivenom (liq-uid, batch 3161299LQ; total protein concentration:6.9 g/dL), obtained by immunization of horses withaa sf loe Gm ac-c -d gas n-t x-p

2

ce.E titu-t CI-C

othrocophias hyoprora(Instituto Nacional de Saluima, personal communication), and neutralizesenoms of the most importantBothropsspecies ineru (Incio-Ruiz et al., 1993). Antivenoms manufac

ured in other Latin American countries have beensed in Peru, such as the polyvalent (Crotalinae)

ivenom produced by Instituto Clodomiro Picadoosta Rica (Zavaleta and Salas, 1996). Despite unublished clinical evidence of effectiveness by this

ivenom in treating snakebite envenomations in Pu,controlled preclinical assessment of the efficac

his antivenom against Peruvian venoms has notublished. This study reports the ability of Costaan polyvalent antivenom, comparatively with polyent ‘bothropic’ antivenom of Peru, to neutralize th

ost relevant toxicological effects induced by the vms ofB. atrox, B. brazili, B. barnettiandB. pictus

rom Peru.

mixture of equal parts of the venoms ofBothropssper, C. durissus durissusandLachesis stenophry

rom Costa Rica (Bolanos and Cerdas, 1980; Angut al., 1997). This antivenom is made of whole Igolecules, purified by caprylic acid precipitation,

ording toRojas et al. (1994). This antivenom is stanardized as to have a neutralizing potency of 3 mB.spervenom, 2 mgC. d. durissusvenom and 3 mgL.tenophrysvenom neutralized per mL antivenom. Aivenoms were stored at 4◦C and used before their eiry time.

.2. Toxic activities of venoms

Toxicological tests were performed in CD-1 mixperimental protocols were approved by the Ins

ional Committee for the Care and Use of Animals (UA) of the University of Costa Rica.

E. Rojas et al. / Acta Tropica 93 (2005) 85–95 87

2.2.1. LethalityLethality was assessed by the intraperitoneal

(i.p.) route. Groups of six mice (16–18 g) were in-jected with various doses of venom, dissolved in0.5 mL of phosphate-buffered saline solution, pH 7.2(PBS). Deaths occurring during 48 h were recorded,and median lethal doses (LD50) were estimatedby the Spearman–Karber procedure (World HealthOrganization, 1981; Gene and Robles, 1987).

2.2.2. Hemorrhagic activityIt was determined by the intradermal test, accord-

ing to Gutierrez et al. (1985). Groups of four mice(18–20 g) were injected intradermally with variousdoses of venom, dissolved in 100�L of PBS. Two hoursafter injection, animals were sacrificed by CO2 inhala-tion, their skins were removed, and the diameter ofthe hemorrhagic area in the inner side of the skin mea-sured. The minimum hemorrhagic dose (MHD) was theamount of venom which induced a hemorrhagic spotof 10 mm diameter (Gutierrez et al., 1985).

2.2.3. Myotoxic activityGroups of four mice (18–20 g) received intramus-

cular (i.m.) injections of various amounts of venom,dissolved in 100�L of PBS, in the right gastrocne-mius muscle (Gutierrez et al., 1980). Control animalswere injected with 100�L PBS. Three hours after in-jection, mice were bled from the orbital plexus, underCO2 anesthesia. Plasma creatine kinase (CK) activityw tiv-i untoc ose( in-d nd-i Sa

2cu-

ta ot-p ofe ow-pi thea in

the thickness of the envenomated foot when comparedwith the PBS-injected foot.

2.2.5. Coagulant activityVarious amounts of venom, dissolved in 100�L of

PBS, were added to 200�L of human citrated plasmaincubated at 37◦C, and the clotting times recorded.The minimum coagulant dose (MCD) was the amountof venom inducing coagulation of plasma in 60 s(Theakston and Reid, 1983; Gene et al., 1989).

2.2.6. Defibrinating activityIt was determined according toTheakston and Reid

(1983), with the modifications ofGene et al. (1989).Groups of four mice (18–20 g) were injected by the in-travenous route, in the tail vein, with various amountsof venom, dissolved in 200�L of PBS. Control micewere injected with 200�L PBS. One hour after injec-tion, mice were bled from the ocular plexus under CO2anesthesia. Immediately, blood was placed in dry glasstubes and kept at room temperature (22–25◦C) for 1 h.Then, tubes were tilted and the formation of a clot ob-served. The minimum defibrinating dose (MMD) wasthe lowest amount of venom that rendered blood un-clottable in all mice injected.

2.3. Immunodetection of myotoxin-likecomponents

The presence of components antigenically related toBf bedps ngp anesa kedw al-b tedw ),w witha BS-B ctedi bu-l ma),u ni-t dev II( s-

as quantitated by the Sigma kit 47-UV. CK acty is expressed in units/L, 1 U defined as the amof enzyme that produces 1�mol of NADH under theonditions of the assay. The minimum myotoxic dMMD) corresponds to the amount of venom thatuces an increment of plasma CK activity correspo

ng to four times the activity of mice injected with PBlone.

.2.4. Edema-forming activityGroups of four mice (18–20 g) were injected sub

aneously, in the right footpad, with 50�L of variousmounts of venom dissolved in PBS. The left foad received 50�L of PBS alone. The thicknessach foot was measured 1 h after injection with a lressure spring caliper (Lomonte et al., 1993). The min-

mum edema-forming dose (MED) corresponds tomount of venom that induces an increment of 30%

. aspermyotoxins in the different venoms ofBothropsrom Peru was determined by dot-blotting, as descrireviously (Lomonte et al., 1990). In brief, venomamples (2�L, containing 1000, 250, 62, 15, or 4rotein) were adsorbed onto nitrocellulose membrnd allowed to dry. The membranes were blocith PBS containing 1% (w/v) of bovine serumumin (BSA), for 2 h. One membrane was incubaith rabbit anti-B. aspermyotoxin II serum (1:500hereas the other served as a control, incubatedn equal amount of normal rabbit serum, for 3 h in PSA. After washing, bound antibodies were dete

n both membranes by an anti-rabbit immunogloin/alkaline phosphatase conjugate (1:1000; Sigsing 5-bromo-4-chloro-3-indolyl phosphate and

roblue tetrazolium for final color development. Cruenom ofB. asper, and a sample of purified myotoxinLomonte and Gutierrez, 1989) were included as po

88 E. Rojas et al. / Acta Tropica 93 (2005) 85–95

itive controls. Reactivity was expressed in semiquan-titative terms, as follows: (−) no reactivity; (+) weakreactivity; (++) strong reactivity.

2.4. Neutralization by antivenoms

Neutralization assays followed the protocol inwhich a constant amount of venom (‘challenge dose’)is incubated, for 30 min at 37◦C, with various dilutionsof antivenom, before testing the corresponding activityin the various pharmacological assay systems describedabove (Gutierrez et al., 1990, 1996). In the case of neu-tralization of lethality, a dose of venom correspondingto four LD50s was utilized, following the methodologyroutinely used at the Quality Control Laboratory of In-stituto Clodomiro Picado (Manual of Quality ControlProcedures, Instituto Clodomiro Picado). The ‘chal-lenge doses’ of venom for the rest of the activitiestested were: (a) hemorrhagic activity: five MHD; (b)myotoxic activity: three MMD; (c) edema-forming ac-tivity: six MED; (d) coagulant activity: two MCD; and(e) defibrinating activity: two MDD.

In the case of lethal, hemorrhagic and edema-forming effects, neutralizing capacity of antivenomswas expressed as effective dose 50%, in two differ-ent ways: (a) the actual volume of antivenom (�L)which reduced by 50% the activity of the ‘challengedose’ of venom, and (b) the ratio of�L antivenom/mg

TT

V ema-foB 5± 1.1B 8± 0.1B 2± 0.1B 3± 0.0

s the d icei

emorrh ah

edema s3 compa

toxic do ni r inject ing PBSa

ulant do go

fibrina sb ation w

ality, w

venom in which the activity was reduced by 50%. Inthe case of coagulant and defibrinating activities, neu-tralization was expressed as effective dose (ED), ac-cording toGene et al. (1989). For coagulant activity,ED corresponds to the volume of antivenom or, al-ternatively, to the ratio�L antivenom/mg venom inwhich the clotting time is prolonged three times whencompared with the clotting time of plasma incubatedwith venom alone. For defibrinating activity, ED cor-responds to the volume of antivenom, or the ratio�Lantivenom/mg venom, in which none of the injectedmice were defibrinated, i.e., all blood samples clotted(Gene et al., 1989).

2.5. Statistical analysis

Results are presented as mean± S.D., with the ex-ception of lethality assays in which variability wasexpressed as 95% confidence limits. The significanceof the differences of the means of two experimentalgroups was determined by the Student’st-test.

3. Results

3.1. Toxicological activities of venoms

Table 1shows the toxicological profiles of the ven-oms studied. The venoms having the highest lethality

able 1oxicological activities ofBothropssp. venoms from Peru

enom Lethala Hemorrhagicb Ed. atrox 6.04 (5.12–7.12)g 6.45± 0.43g 2.9. brazili 5.51 (4.76–6.37) 3.71± 0.41 1.1. pictus 2.78 (2.31–3.34) 1.06± 0.15 0.5. barnetti 2.66 (2.30–3.07) 0.75± 0.26 1.1

a Lethality is expressed as median lethal dose (LD50), defined anjected intraperitoneally.

b Hemorrhagic activity is expressed in terms of minimum hemorrhagic spot of 10 mm diameter 2 h after injection.

c Edema-forming activity is expressed in terms of minimum0% increment in the thickness of envenomated footpad when

d Myotoxic activity is expressed in terms of minimum myoncrement in plasma CK activity in mice 3 h after intramusculalone.

e Coagulant activity is expressed in terms of minimum coagf human plasma in 60 s.

f Defibrinating activity is expressed in terms of minimum delood incoagulability 1 h after injection in all mice. ND: defibrin

g Results of are presented as mean± S.D. (n= 4), except in leth

rmingc Myotoxicd Coagulante Defibrinatingf

1g 56.3± 9.93g 3.65± 0.02g 2.5g

4 13.5± 0.54 7.39± 0.06 ND3 15.4± 1.49 46.6± 0.40 105 40.4± 1.55 21.5± 0.45 2.5

ose of venom (�g venom per g body weight) that kills 50% of m

agic dose (MHD), defined as the amount of venom (�g) that induces

-forming dose (MED), defined as the amount of venom (�g) that inducered with non-envenomated contralateral footpad.se (MMD), defined as the amount of venom (�g) which induces a

ion corresponding to four times the CK activity of mice receiv

se (MCD), defined as the amount of venom (�g) which induces clottin

ting dose (MDD), defined as the amount of venom (�g) which induceas not observed at the highest sublethal dose tested (5�g).here 95% confidence limits are included in parentheses.

E. Rojas et al. / Acta Tropica 93 (2005) 85–95 89

Table 2Neutralization of lethality of PeruvianBothropssp. venoms by antivenoms produced in Peru and Costa Rica,Neutralization (effective dose 50%)a

Venom �L antivenom/four LD50s �L antivenom/mg venom mg venom/mL antivenom

Peruvianantivenom

Costa RicanAntivenom

Peruvianantivenom

Costa RicanAntivenom

Peruvianantivenom

Costa Ricanantivenom

B. atrox* 83 (60–114) 186 (135–257) 201 (145–277) 452 (328–625) 4.98 (3.61–6.88) 2.21 (1.60–3.05)B. brazili 138 (96–199) 138 (96–199) 369 (256–532) 369 (256–532) 2.71 (1.88–3.91) 2.71 (1.88–3.91)B. pictus 57 (41–79) 94 (67–133) 301 (218–417) 500 (356–704) 3.32 (2.40–4.59) 2.00 (1.42–2.81)B. barnetti* 67 (48–92) 185 (127–266) 369 (267–510) 1020 (704–1470) 2.71 (1.96–3.74) 0.98 (0.68–1.42)

∗ Significantly different between antivenoms.a Neutralization of lethality is expressed in three different ways to facilitate comparisons: (i)�L antivenom per challenge dose of venom

(four LD50s), (ii) �L antivenom per mg venom, an expression used in previous studies, and (iii) mg venom neutralized per mL antivenom, anexpression typically used in Production Laboratories and included in the prospects of these antivenoms. Values in parentheses represent the 50%confidence limits.

wereB. barnettiandB. pictus, followed by that ofB.brazili, whereas the venom ofB. atroxhad the lowesttoxicity. All venoms induced local tissue damage, i.e.,hemorrhage, edema and myonecrosis, with significantdifferences between their relative activities (Table 1).The venom ofB. atrox showed a relatively low my-otoxic activity. Immunodetection of components thatcross-react withB. aspermyotoxin II revealed posi-tive results in all venoms tested; the strongest reactiv-ity (++) was present in the venom ofB. brazili, withweaker reactions (+) observed in the rest of the venoms.All venoms presented coagulant activity, and those ofB. atrox,B. pictusandB. barnettishowed defibrinatingactivity. In the case ofB. brazili venom, defibrinationwas not observed with the highest dose tested (5�g);higher doses resulted in lethality before 1 h, which isthe sampling time for the assessment of defibrination.

3.2. Neutralization studies

Both antivenoms were effective in the neutraliza-tion of lethality induced by all venoms studied. Pe-ruvian antivenom showed higher efficacy against thevenoms ofB. atrox andB. barnetti, whereas a sim-ilar neutralization was achieved with the venoms ofB. pictusandB. brazili (Table 2). In order to deter-mine if the differences in efficacy between antivenomsare due to antigenic variations in the venoms used asimmunogens or to a higher general potency of Peru-v u-t aR lizedt d a

higher potency [ED50 of 322�L antivenom/mg venom(95% confidence interval: 282–369�L/mg)] than Peru-vian antivenom [ED50 of 500�L antivenom/mg venom(95% confidence interval: 356–599�L/mg)]. Neutral-ization of hemorrhagic activity of Peruvian venomswas quite effective with both antivenoms, being high-est with the Peruvian ‘bothropic’ antivenom (Table 3).Edema-forming effect was neutralized by both antiven-oms, with quantitative variations depending on thevenom tested (Table 3). In the case of neutralizationof B. brazili venom by Peruvian antivenom, the neu-tralization curve was atypical, since an increment inedema-forming activity was observed when high an-tivenom doses were used, and no 50% neutralizationwas achieved (Fig. 1A), whereas the curve correspond-ing to Costa Rican antivenom showed a characteristicdose-dependent neutralization of this activity (Fig. 1B).

Regarding myotoxicity, both antivenoms effectivelyneutralized this activity of the venoms ofB. brazili, B.pictusandB. barnetti(Table 3). Costa Rican antivenomhad higher efficacy in the neutralization of this activityof B. brazilivenom. Neutralization of myotoxicity wasnot studied in the case ofB. atroxvenom due to the highvalue of the MMD; a ‘challenge dose’ of three MMDscaused a rather severe local tissue reaction (hemorrhageand edema), which was avoided. Both antivenoms neu-tralized coagulant activity of the venoms ofB. brazili,B. atroxandB. pictus(Table 4). Costa Rican antivenomeffectively neutralized coagulant effect ofB. barnettivenom (Fig. 2B); however, in the case of Peruvian an-t r, ana thisc an-

ian antivenom againstBothropssp. venoms, the neralization of lethality ofB. aspervenom from Costica was also assessed. Both antivenoms neutra

his effect, but the Costa Rican antivenom showe

ivenom, no neutralization was achieved. Moreovetypical neutralization curve was again observed inase, with reduction in the clotting times at some

90 E. Rojas et al. / Acta Tropica 93 (2005) 85–95

Table 3Neutralization of hemorrhagic, edema-forming and myotoxic activities of PeruvianBothropssp. venoms by antivenoms produced in Peru andCosta Rica

Venom Peruvian antivenom Costa Rican antivenom Peruvian antivenom Costa Rican antivenom

Neutralization of hemorrhagic activity (ED50)a

�L antivenom/five MHD �L antivenom/mg venom

B. atrox* 4.45± 0.36 8.24± 1.21 163± 13 302± 44B. brazili* 0.78± 0.01 2.55± 0.17 42± 1 138± 9B. pictus* 0.86± 0.08 2.13± 0.53 163± 14 403± 99B. barnetti* 0.58± 0.09 1.13± 0.23 155± 25 302± 61

Neutralization of edema-forming activity (ED50)a

�L antivenom/six MED �L antivenom/mg venom

B. atrox* 17.6± 4.3 8.1± 1.9 995± 243 455± 107B. brazili* NNb 10.3± 1.4 NNb 1452± 203B. pictus* 0.60± 0.17 1.37± 0.23 194± 54 440± 75B. barnetti 2.13± 0.80 1.83± 0.55 314± 118 270± 81

Neutralization of myotoxic activity (ED50)a

�L antivenom/three MMD �L antivenom/mg venom

B. brazili* 14.0± 0.6 8.1± 0.6 358± 16 200± 14B. pictus 9.0 ± 3.1 6.4± 1.1 196± 67 139± 25B. barnetti 34.4± 3.9 38.2± 3.4 284± 32 315± 28

∗ The differences between antivenoms are statistically significant (P< 0.05).a Neutralization of hemorrhagic, edema-forming and myotoxic activities is expressed as effective dose 50%, in two different ways: (i)�L

of antivenom per challenge dose of venom (five minimum hemorrhagic doses, six minimum edema-forming doses or three minimum myotoxicdoses) and (ii)�L antivenom per mg venom which reduce by 50% the effect induced by venom alone (see Section2). Results are presented asmean± S.D. (n= 4).

b NN: no 50% neutralization of the effect was achieved at the highest antivenom/venom level tested (2000�L antivenom/mg venom).

Table 4Neutralization of coagulant and defibrinating activities of PeruvianBothropssp. venoms by antivenoms produced in Peru and Costa Rica

Venom Peruvianantivenom Costa Rican antivenomPeruvianantivenom Costa Rican antivenom

Neutralization of coagulant activity (ED)a

�L antivenom/two MCD �L antivenom/mg venom

B. atrox 1.1 ± 0.003 1.1± 0.003 151± 0.5 150± 0.4B. brazili* 4.0 ± 0.003 0.96± 0.013 273± 0.2 65± 0.9B. pictus* 7.33± 0.07 11.65± 0.14 79± 1 125± 2B. barnetti* NNb 6.70± 0.29 NNb 156 ± 7

Neutralization of defibrinating activity (ED)a

�L antivenom/two MDD �L antivenom/mg venom

B. atrox* 5 2.5 1000 500B. pictus 10 10 500 500B. barnetti 2.5 2.5 500 500

∗ The differences between antivenoms are statistically significant (P< 0.05).a Neutralization of coagulant and defibrinating activities is expressed as effective dose in two different ways: (i)�L of antivenom per challenge

dose of venom (two minimum coagulant doses or two minimum defibrinating doses) and (ii)�L antivenom per mg venom which reduce by50% the effect induced by venom alone (see Section2). Results are presented as mean± S.D. (n= 4).

b NN: no 50% neutralization of the effect was achieved at the highest antivenom/venom level tested (2000�L antivenom/mg venom).

E. Rojas et al. / Acta Tropica 93 (2005) 85–95 91

Fig. 1. Neutralization of edema-forming activity ofB. brazilivenomby antivenoms of Peru (A) and Costa Rica (B). Mixtures containinga constant amount of venom and various dilutions of antivenom wereincubated at 37◦C for 30 min; afterwards, aliquots of the mixtures(50�L containing six minimum edema-forming doses of venom)were injected in the right footpad of mice, whereas the left foot padreceived 50�L of PBS. The thickness of both feet was estimatedone hr after injection with a low-pressure spring caliper, and edemawas expressed as the percentage increment in thickness of the rightfootpad as compared to the left one. A typical dose-dependent neu-tralization is observed in (B), whereas in (A) edema was reducedat the lowest antivenom/venom dose, increasing at higher doses ofantivenom.

Fig. 2. Neutralization of coagulant activity ofB. barnetti venomby antivenoms of Peru (A) and Costa Rica (B). Mixtures contain-ing a constant amount of venom and various dilutions of antivenomwere incubated at 37◦C for 30 min; afterwards, aliquots of the mix-tures (100�L containing two minimum coagulant doses of venom)were added to 200�L of human citrated plasma, and clotting timesrecorded. A typical neutralization curve is observed in (B), withprolongation of clotting times as antivenom doses increased; at an-tivenom/venom ratios of 500 and 1000�L/mg plasma did not clotduring an observation time of 5 min. In contrast, no neutralizationwas observed in (A), since clotting times were not prolonged af-ter incubation with antivenom, and were even reduced at the lowerantivenom/venom ratios tested.

92 E. Rojas et al. / Acta Tropica 93 (2005) 85–95

tivenom/venom ratios (Fig. 2A). A procoagulant activ-ity of antivenom was discarded, since addition of vari-ous volumes of antivenom alone to citrated plasma didnot result in clotting. Defibrinating activity was neutral-ized by both antivenoms for all venoms tested (Table 4).

4. Discussion

The four venoms studied present a toxicologicalprofile similar to those of otherBothrops sp. ven-oms in Latin America. This profile is characterizedby prominent local tissue alterations, i.e., hemorrhage,myonecrosis and edema, and by systemic manifesta-tions associated with hemorrhage and coagulopathy(Fan and Cardoso, 1995; Gutierrez, 1995). Our obser-vations evidenced that polyvalent ‘antibothropic’ an-tivenom and polyvalent (Crotalinae) antivenom, pro-duced in Peru and Costa Rica, respectively, are effec-tive in the neutralization of the most important toxicactivities induced by these venoms when using an ex-perimental protocol based on incubation of venom andantivenom prior to testing in the experimental systems(Gutierrez et al., 1990, 1996). Thus, despite the fact thatvenoms from Peruvian species are not included in theimmunizing mixture used in Costa Rica, the antivenomproduced from the plasma of horses immunized withthe venoms ofB. asper, C. durissus durissusandL.stenophrysis effective against these fourBothropsven-oms from Peru. These preclinical observations agreew an-t ionsi -s n-o iffer-e atings -s izedc

ntss ationo icals omso l-i re-s e ofe r-t ted,

whereas edema increased at higher antivenom levels.This phenomenon was not observed with the CostaRican antivenom, where a dose-dependent reductionin edema-forming effect occurred. Moreover, Peruvianantivenom did not neutralize coagulant activity ofB.barnettivenom; instead, coagulant activity was morepronounced at some antivenom/venom ratios. How-ever, Peruvian antivenom was effective in the neutral-ization of defibrinating activity of all venoms tested.Since defibrination is mainly due to the action of pro-coagulant components acting in vivo, it is suggestedthat both antivenoms are effective in neutralizing tox-ins responsible for coagulopathy in vivo, despite theanomalous neutralization by Peruvian antivenom of co-agulant activity ofB. barnettiobserved in vitro.

A similar phenomenon of reduced neutralization athigh antivenom/venom ratios had been described for aformer polyvalent antivenom produced in Costa Ricaby ammonium sulfate precipitation of immunoglob-ulins (Gutierrez et al., 1986; Gene et al., 1989). Itwas suggested that such effect is due to the presencein the antivenom of non-IgG proteins from equineplasma that, in contact with venom enzymes, may re-lease procoagulant (Gene et al., 1989) or proinflam-matory (Gutierrez et al., 1986) proteins or peptidesduring the 30 min incubation time. Therefore, whenthe venom/antivenom mixture is injected into mice oradded to plasma, these substances would cause edemaor clotting, respectively. Since the Peruvian antivenomused in this study was produced by ammonium sulfatef amep trast,t byc ro-c ,1

tesi s,1S int liza-t lo-c oxicp int in-i f-f oo ases,

ith earlier reports of the efficacy of Costa Ricanivenom in the treatment of snakebite envenomatn Peru (Zavaleta and Salas, 1996). The differences oberved in the values of ED50s between these antivems, for some of the effects studied, suggest that dnt dosage protocols may have to be used when trenakebites with these antivenoms in Peru, a hypotheis that needs to be tested in controlled, randomlinical trials.

Most neutralization curves in our experimehowed a characteristic dose-dependent neutralizf the toxic activities tested. However, an atyphape in these curves was observed with the venf B. brazili andB. barnettiwhen studying neutra

zation of edema-forming and coagulant activities,pectively, by the Peruvian antivenom. In the casdema induced byB. brazili venom, there was pa

ial neutralization at the lowest antivenom dose tes

ractionation of plasma, it is suggested that the shenomenon occurred in the present study. In con

he Costa Rican antivenom is currently producedaprylic acid precipitation of non-IgG proteins, a pedure that yields a more purified product (Rojas et al.994).B. atroxis responsible for the majority of snakebi

n the Amazonian region of Peru (Zavaleta and Sala996) and other South American countries (dosantos, 2003); therefore, antivenoms to be used

hese regions have to be effective in the neutraion of this venom. This venom induces prominental tissue damage, although the content of myothospholipases A2 was low in the venom used

his study, in agreement with its relatively high mmum myotoxic dose. Thus, it is likely that local eects induced by PeruvianB. atroxare mainly due tther venom components, such as metalloprotein

E. Rojas et al. / Acta Tropica 93 (2005) 85–95 93

known to cause severe local pathological alterations(Gutierrez and Rucavado, 2000). Both antivenoms neu-tralized toxic activities ofB. atrox venom. Peruvianantivenom was more effective in the neutralization oflethal and hemorrhagic activities, whereas Costa Ricanantivenom was more effective against defibrinating andedema-forming activities, having a similar effective-ness in the neutralization of coagulant activity. It is con-cluded that extensive immunological cross-reactivityoccurs between components present in Central Amer-icanB. asperand South AmericanB. atrox. This con-clusion is reinforced by the observation that Peruvianantivenom is effective in the neutralization of lethal-ity induced byB. aspervenom from Costa Rica, albeithaving a lower efficacy than that of Costa Rican poly-valent antivenom. Therefore, quantitative variations inED50s between antivenoms are due to antigenic dif-ferences between toxins present in the venoms, in thecontext of a significant cross-reactivity between theseantigens.

An extensive cross-neutralization between severalBothropsvenoms and antivenoms produced in vari-ous Latin American countries has been previously de-scribed (Gutierrez et al., 1985; Rojas et al., 1987; Diasda Silva et al., 1989; Otero et al., 1995; Theakstonet al., 1995; de Roodt et al., 1998; Bogarın et al., 2000;Saravia et al., 2001), revealing important immunolog-ical similarities between relevant toxins inBothropsvenoms from different countries. These preclinical ob-

iven-therlo-hastri-zil

s-rolo-herd inheress-

an-ilrid

used

tional collaborative studies are required in Latin Amer-ica to further characterize the neutralizing spectrumof antivenoms produced in this region against a vari-ety of snake venoms from Central and South America.Moreover, these preclinical studies must be followedby randomized controlled clinical trials as the final cri-terion for the efficacy and safety of antivenoms in theregion.

Acknowledgements

The authors thank Ana Marıa Espinoza and the staffof Centro Nacional de Produccion de Biologicos, In-stituto Nacional de Salud, Peru, for providing the ven-oms and the Peruvian antivenom, as well as SteveQuiros, Teresa Escalante and Javier Nunez, from In-stituto Clodomiro Picado, for their collaboration. Thisinvestigation was supported by Vicerrectorıa de Inves-tigacion, Universidad de Costa Rica (project 741-A1-027).

References

Angulo, Y., Estrada, R., Gutierrez, J.M., 1997. Clinical and labora-tory alterations in horses during immunization with snake ven-oms for the production of polyvalent (Crotalinae) antivenom.Toxicon 35, 81–90.

Bogarın, G., Segura, E., Duran, G., Lomonte, B., Rojas, G., Gutierrez,nos

B , M.,a

-

B .A.,Hi-ineoms.

B n-

C atin

d E.,cro-8.

D .P.,K.,

J.M., 1995. Evaluacion de la capacidad de cuatro antivenecomerciales para neutralizar el veneno de la serpienteBothropsasper(terciopelo) de Costa Rica. Toxicon 33, 1242–1247.

ogarın, G., Romero, M., Rojas, G., Lutsch, C., CasadamontLang, J., Otero, R., Gutierrez, J.M., 1999. Neutralization, bymonospecificBothrops lanceolatusantivenom, of toxic activities induced by homologous and heterologousBothropssnakevenoms. Toxicon 37, 551–557.

ogarın, G., Morais, J.F., Yamaguchi, I.K., Stephano, MMarcelino, J.R., Nishikawa, A.K., Guidolin, R., Rojas, G.,gashi, H.G., Gutierrez, J.M., 2000. Neutralization of crotalsnake venoms from Central and South America by antivenproduced in Brazil and Costa Rica. Toxicon 38, 1429–1441

olanos, R., Cerdas, L., 1980. Produccion y control de sueros atiofıdicos en Costa Rica. Bol. Sanit. Panam. 88, 189–196.

ampbell, J.A., Lamar, W., 1989. The Venomous Reptiles of LAmerica. Cornell University Press, Ithaca, NY, p. 425.

e Roodt, A.R., Dolab, J.A., Fernandez, T., Segre, L., Hajos, S.1998. Cross-reactivity and heterologous neutralization oftaline antivenoms used in Argentina. Toxicon 36, 1025–103

ias da Silva, W., Guidolin, R., Raw, I., Higashi, H.G., Caricati, CMorais, J.F., Lima, M.L.S.R., Yamaguchi, I.K., Nishikawa, A.

servations support the hypothesis that some antoms are effective against heterologous venoms in ocountries, and may be helpful in situations in whichcally produced antivenoms are not available. Thisbeen properly demonstrated in controlled clinicalals in Colombia with antivenoms produced in Braand Costa Rica (Otero et al., 1996, 1999; Otero-Patinoet al., 1998). However, it is important to carefully asess the capacity of antivenoms to neutralize hetegous venoms before their clinical application in otcountries, with methodologies such as those usethis work, since there have been reports of cases wsome antivenoms have proven ineffective in croneutralization studies (Theakston et al., 1995; Bogarınet al., 1995, 1999). A recent study demonstrated thattivenoms manufactured in Peru, Colombia and Brazare effective in the neutralization of Peruvian vipesnake venoms in similar preclinical tests as thosein the present investigation (Laing et al., 2004). Addi-

94 E. Rojas et al. / Acta Tropica 93 (2005) 85–95

Stephano, M.A., Marcelino, J.R., Pinto, J.R., Santos, M.J., 1989.Cross-reactivity of horse monovalent antivenoms to venoms oftenBothropsspecies. Mem. Inst. Butantan. 51, 153–168.

dos Santos M.C., 2003. Serpentes peconhentas e ofidismo no Ama-zonas. In: J.L.C. Cardoso, F.O.S. Franca, F.H. Wen, C.M.S.Malaque, V. Hadad (Eds), Animais Peconhentos no Brasil, SaoPaulo, pp. 115–125.

Fan, H.W., Cardoso, J.L., 1995. Clinical toxicology of snake bitesin South America. In: Meier, J., White, J. (Eds.), Handbook ofClinical Toxicology of Animal Venoms and Poisons. CRC Press,Boca Raton, FL, pp. 667–688.

Gene, J.A., Robles, A., 1987. Determinacion de la dosis letal 50%por el metodo de Spearman-Karber. Revista Medica HospitalNacional de Ninos (Costa Rica) 22, 25–40.

Gene, J.A., Roy, A., Rojas, G., Gutierrez, J.M., Cerdas, L., 1989.Comparative study on the coagulant, defibrinating, fibrinolyticand fibrinogenolytic activities of Costa Rican crotaline snakevenoms and their neutralization by a polyvalent antivenom. Tox-icon 27, 841–848.

Gutierrez, J.M., 1995. Clinical toxicology of snake bite in CentralAmerica. In: Meier, J., White, J. (Eds.), Handbook of ClinicalToxicology of Animal Venoms and Poisons. CRC Press, BocaRaton, FL, pp. 645–665.

Gutierrez, J.M., Rucavado, A., 2000. Snake venom metallopro-teinases: their role in the pathogenesis of local tissue damage.Biochimie 82, 841–850.

Gutierrez, J.M., Arroyo, O., Bolanos, R., 1980. Mionecrosis, hemor-ragia y edema inducidos por el veneno deBothrops asperen ratonblanco. Toxicon 18, 603–610.

Gutierrez, J.M., Gene, J.A., Rojas, G., Cerdas, L., 1985. Neutral-ization of proteolytic and hemorrhagic activities of Costa Ricansnake venoms by a polyvalent antivenom. Toxicon 23, 887–893.

Gutierrez, J.M., Rojas, G., Lomonte, B., Gene, J.A., Cerdas, L., 1986.Comparative study of the edema-forming activity of Costa Ricansnake venoms and its neutralization by a polyvalent antivenom.

G l-esting29.

G a-t ofoyf-ation

I s-

L .,mer-rtant

L xin,

L jas,ake

venom proteins antigenically-related toBothrops aspermyotox-ins. Braz. J. Med. Biol. Res. 23, 427–435.

Lomonte, B., Tarkowski, A., Hanson, L.A, 1993. Host response toBothrops aspersnake venom: Analysis of edema formation, in-flammatory cells and cytokine release in a mouse model. Inflam-mation 17, 93–105.

Organizacion Panamericana de la Salud 1977. Manual de Proced-imientos. Produccion y Pruebas de Control en la Preparacionde Antisueros Difterico, Tetanico, Botulınico, Antivenenos y dela Gangrena Gaseosa. Organizacion Panamericana de la Salud,Washington, DC.

Otero, R., Nunez, V., Osorio, R.G., Gutierrez, J.M., Giraldo, C.A.,Posada, L.E., 1995. Ability of six Latin American antiven-oms to neutralize the venom of Mapana Equis (Bothrops atrox)from Antioquia and Choco (Colombia). Toxicon 33, 809–815.

Otero, R., Gutierrez, J.M., Nunez, V., Robles, A., Estrada, R., Segura,E., Toro, M.F., Garcıa, M.E., Dıaz, A., Ramırez, E.C., Gomez,G., Castaneda, J., Moreno, M.E., 1996. A randomized doubleblind clinical trial of two antivenoms in patients bitten byBoth-rops atroxin Colombia. Trans. Royal Soc. Trop. Med. Hyg. 90,696–700.

Otero, R., Gutierrez, J.M., Rojas, G., Nunez, V., Dıaz, A., Miranda,E., Uribe, A.F., Silva, J.F., Ospina, J.G., Medina, Y., Toro, M.F.,Garcıa, M.E., Leon, G., Garcıa, M., Lizano, S., De la Torre, J.,Marquez, J., Mena, Y., Gonzalez, N., Arenas, L.C., Puzon, A.,Blanco, N., Sierra, A., Espinal, M.E., Arboleda, M., Jimenez,J.C., Ramırez, P., Dıaz, M., Guzman, M.C., Barros, J., Henao,S., Ramırez, A., Macea, U., Lozano, R., 1999. A randomizedblinded clinical trial of two antivenoms, prepared by caprylicacid or ammonium sulphate fractionation of IgG, inBothropsandPorthidium snake bites in Colombia: correlation betweensafety and biochemical characteristics of antivenoms. Toxicon37, 895–908.

Otero-Patino, R., Cardoso, J.L.C., Higashi, H.G., Nunez, V., Sierra,-,ara-omsd.

R 7.ontive-

. 35,

R c-ple

S ,ofza-

T , A.,ster,.a

Comp. Biochem. Physiol. 85C, 171–175.utierrez, J.M., Rojas, G., Lomonte, B., Gene, J.A., Chaves, F., A

varado, J., Rojas, E., 1990. Standardization of assays for tthe neutralizing ability of antivenoms. Toxicon 28, 1127–11

utierrez, J.M., Rojas, G., Bogarın, G., Lomonte, B., 1996. Evalution of the neutralizing ability of antivenoms for the treatmensnake bite envenoming in Central America. In: Bon, C., Gfon, M. (Eds.), Envenomings and Their Treatments. FondMarcel Merieux, Lyon, pp. 223–231.

ncio-Ruiz, R., Incio-Ruiz, L., Zavaleta Martınez-Vargas, A., SalaArruz, M., Gutierrez, J.M., 1993. Toxicidad y neutralizacion devenenos ofıdicos peruanos de los generosBothropsy Lachesis(Serpentes: Viperidae). Rev. Biol. Trop. 41, 351–357.

aing, G.D., Yarleque, A., Marcelo, A., Rodrıguez, E., Warrell, D.ATheakston, R.D.G., 2004. Preclinical testing of three south Aican antivenoms against the venom of five medically-impoPeruvian snake venoms. Toxicon. 44, 103–106.

omonte, B., Gutierrez, J.M., 1989. A new muscle damaging tomyotoxin II, from the venom of the snakeBothrops asper(ter-ciopelo). Toxicon 27, 725–733.

omonte, B., Furtado, M.F., Rovira, M.E., Carmona, E., RoG., Aymerich, R., Gutierrez, J.M., 1990. South American sn

A., Dıaz, A., Toro, M.F., Garcıa, M.E., Moreno, A.M., Medina, M.C., Castaneda, N., Silva-Dıaz, J.F., Murcia, M., CardenasS.Y., Dias da Silva, W., 1998. A randomized, blinded, comptive trial of one pepsin-digested and two whole IgG antivenfor Bothropssnake bites in Uraba Colombia. Am. J. Trop. MeHyg. 58, 183–189.

ojas, G., Gutierrez, J.M., Gene, J.A., Gomez, M., Cerdas, L., 198Neutralizacion de las actividades toxicas y enzimaticas de cuatrvenenos de serpientes de Guatemala y Honduras por el aneno polivalente producido en Costa Rica. Rev. Biol. Trop59–67.

ojas, G., Jimenez, J.M., Gutierrez, J.M., 1994. Caprylic acid frationation of hyperimmune horse plasma: description of a simprocedure for antivenom production. Toxicon 32, 351–363.

aravia, P., Rojas, E., Escalante, T., Arce, V., Chaves, E., VelasquezR., Lomonte, B., Rojas, G., Gutierrez, J.M., 2001. The venomBothrops asperfrom Guatemala: toxic activities and neutralition by antivenoms. Toxicon 39, 401–405.

heakston, R.D.G., Laing, G.D., Fielding, C.M., Freire LascanoTouzet, J.M., Vallejo, F., Guderian, R.H., Nelson, S.J., WuW., Richards, A.M., Rumbea-Guzman, J., Warrell, D.A., 1995Treatment of snake bites byBothropsspecies andLachesis mut

E. Rojas et al. / Acta Tropica 93 (2005) 85–95 95

in Ecuador: laboratory screening of candidate antivenoms. Trans.Royal Soc. Trop. Med. Hyg. 89, 550–554.

Theakston, R.D.G., Reid, H.A., 1983. Development of simple stan-dard assay procedures for the characterization of snake venoms.Bull. World Health Org. 61, 949–956.

Theakston, R.D.G., Warrell, D.A., 1991. Antivenoms: a list of hyper-immune sera currently available for the treatment of envenomingby bites and stings. Toxicon. 29, 1419–1470.

World Health Organization, 1981. Progress in the Characterizationof Venoms and Standardization of Antivenoms, 58. WHO OffsetPublication, Geneva.

Zavaleta, A., Salas, M., 1996. Ofidismo: envenenamientopor mordedura de serpientes. In: Martınez-Villaverde, J.R.,Leon-Barua, R., Vidal-Neira, L., Losno-Garcıa, R. (Eds.),Emergencias en Medicina Interna. Lima, Peru, pp. 241–260.