Number of vector bites determining the infection of guinea pigs with Trypanosoma cruzi

EPIDEMIOLOGY

RESEARCH NOTE

Number of vector bites determining the infection of guinea pigs with Trypanosoma cruzi

Miguel Angel BasombríoI, ⁺; David Gorla^II; Silvia Catalá^II; María A Segura^I; María C Mora^I; Laura Gómez^I; Julio Nasser^I

^ILaboratorio de Patología Experimental, Facultad de Ciencias de la Salud, Universidad Nacional de Salta, Calle Buenos Aires 177, 4400 Salta, Argentina

^IIFacultad de Ciencias Exactas, Físicas y Naturales, Universidad Nacional de Córdoba, Córdoba, Argentina

Key words: guinea pigs - Trypanosoma cruzi - bites - infection

The number of bites from Trypanosoma cruzi infected Triatoma infestans received by a mammal is one of the main determinants of its risk of becoming infected by this parasite. Although T. cruzi is not introduced by the bite itself, it is mostly during the bite that T. cruzi contaminated feces are deposited on the skin (C Chagas 1909 Mem Inst Oswaldo Cruz 1: 159-218, E Brumpt 1912 Bull Soc Path Exot 5: 723-724).

Domestic guinea pig corrals or "cuyeras" are known as foci of T. cruzi spread (RA Torrico 1950 Bol of Sanit Panam 29: 827-840) and have also been used for experimental purposes (MA Basombrío et al. 1987 Am J Trop Med Hyg 37: 57-62). When guinea pigs are placed in T. infestans-colonized corrals, the number of bites necessary for infecting a guinea pig (NBNI) should be equal to the number of bites taking place in the corral during the time of exposure (A), divided by the number of infected guinea pigs (B):

Since often not all guinea pigs are infected, factor B can be expressed as the number of animals in the corral (N) multiplied by the proportion of animals infected (I), so that:

After the first animal becomes infected, some of the bites are "wasted" on infected animals and should be substracted from NBNI. Since I increases with time, these bites are estimated as one half of those received by the animals presenting infection at the end of the exposure period. This total number of redundant bites is then divided by N, to estimate those corresponding to one guinea pig, so that

and using a common denominator:

Regarding factor A, several measurements can allow its estimation with increased (although not complete) accuracy. Taking into account the number of insects in the corral at the time of censing (n), the number of days vectors and hosts have lived together (d) and the daily proportion of fed vectors (PFV), factor A could be estimated as n x PFV x d. Furthermore, not all bugs are infected and it is possible to examine them and determine the proportion of vectors infected (PVI). Thus, it was possible to dissect formula 2, incorporating all the determinants just mentioned:

Measurements for each of these factors could be obtained in the field post of our laboratory. After an initial, pilot determination previously reported within an ecologic study (S Catalá et al. 1992 Am J Trop Med Hyg 47: 20-26), we present here the results of 11 sets of data from 4 separate experiments performed in uniform, comparable conditions. The post is placed in Cobos (24'47 S 65'06 W), Province of Salta, Argentina, a Chagas'disease endemic rural area, now under insecticide control. Average temperatures range from 9°C in winter to 30° C in summer and humidity from 60 to 100%. A system based on standardized guinea pig corrals, made of loose brick, measuring 1 m² x 40 cm height, and isolated by two layers of mosquito nets (Fig.) was used.

In experiments 1, 2 and 3, five guinea pigs and an original seeding population 696 T. cruzi bearing T. infestans bugs were initially placed in a corral. During 4 to 7 months, this population was censused at intervals, but care was taken not to introduce changes in either bugs or guinea pigs. The original bug seed consisted of 42 adult (12 male and 30 female) and 654 nymphal stages (57N5, 72N4, 290N3, 191N2 and 44N1). In experiment 4, the number of guinea pigs was eight and the vector population started with 696, mostly adult, non stratified T. infestans bugs. In all experiments, initial PVI exceeded 95% in samples and was assumed to be 100% in populations.

n: the number of bugs was determined by periodic censuses. The yards were completely disassembled and all bugs were collected and taken to the laboratory. Developmental stages and sexes were recorded but only the total number of bugs was considered in this work.

PFV: the method used to estimate this factor (S Catalá et al. 1991 Med Vet Entomol 5: 325-333) was based on the finding of transparent urine in the rectal ampoule of bugs fed during the last 24 hr. Since PFV is temperature-dependent (Catalá et al. loc. cit.), the average temperature recorded during each period between censuses was used to adjust PFV in the following census.

PVI: feces from bugs were examined under the microscope for the presence of T. cruzi. For both PFV and PVI determinations, samples of 140 bugs (20 for each developmental stage) were used.

N and I: the number and proportion of guinea pigs carrying T. cruzi infection was recorded. Infection was determined by the microhematocrit method (H Freilij et al. 1986 J Clin Microbiol 18: 327-330) using three capillaries per animal, followed by xenodiagnosis with 20 nymphs and a serologic test (direct agglutination or ELISA).

The Table presents data of 11 censuses from 4 separate experiments. It can be seen that, in 3 censuses (No. 1, 4 and 7), the number of bites (<519) was too small to produce any infection (I=0). In census No. 11, the number of bites was 13624, a number equal or higher than necessary to infect all animals in the yard (I=1). In 7 censes, the number of bites that produced the infection of one guinea pig fell between 1650 and 10957 ( = 5222; SE = 1402). This range is slightly higher, but not substantially distinct from a previous, preliminary determination by our group for a larger, partially sampled guinea pig corral (NBNI = 1461; Catalá et al. loc. cit.) or to estimates for human infection by T. cruzi in triatomine infested human dwellings (NBNI = 1000 to 2500, JE Rabinovich et al. 1990 Bull WHO 68: 737-746).

Some possible confounding factors have to be considered, which possibly reduce the accuracy of these estimates. Underestimation of NBNI may stem from infections not related to bites, such as may occur by contamination of drinking water or food with T. infestans feces or by direct ingestion of infected bugs by the guinea pigs. This alternative was suspected in this system by a rare episode of sudden infection of all animals in a corral with very high density of bugs and an open drinking water dish. In this study, we have attempted to reduce this possibility by using improved water dispensers (Fig.) and avoiding an excessive number of bugs. Insects failing to collect transparent urine after unsuccessful bites may also result in underestimation of NBNI. On the other hand, overestimation of NBNI can result from bug populations with an excess of younger developmental stages with low vectorial capacity, such as occurs early in the summer. Finally, a crucial and highly variable factor in transmission, e.g., the concentration of infective, metacyclic T. cruzi stages in bug's feces, has not been considered as such in these estimates. Bug infectivity has instead been approached in this study by PVI, a parameter which is less influential in transmission. In spite of these possible inaccuracies, this system allows the measurement of factors which decisively correlate the behaviour of T. infestans and the transmission of T. cruzi.

Acknowledgment: to the reviewer of this manuscript and Lic. Eusebio Cleto del Rey for their helpful discussion and contribution.

Received 24 July 1995

Accepted 8 May 1996

This research was supported by the UNDP/ Workd Bank/WHO Special Programme for Research and Training in Tropical Diseases, by CONICET (Consejo Nacional de Investigaciones Científicas y Técnicas), and by Consejo de Investigación, Universidad Nacional de Salta.

+

Corresponding author. Fax: [phone]

Publication Dates

• Publication in this collection
  29 Sept 2008
• Date of issue
  Aug 1996