Detection of anti-Giardia lamblia serum antibody among children of day care centers Detecção de anticorpos séricos anti-Giardia lamblia em crianças de creches

Correspondence to: Semíramis Guimarães Departamwento de Parasitologia Instituto de Biociências da Unesp 18618-000 Distrito de Rubião Júnior, SP, Brasil E-mail: sgviana@ibb.unesp.br Subvencionado pela Fundação da Universidade Estadual de São Paulo (Fundunesp – Processo no 387/91). Edição subvencionada pela Fundação de Amparo à Pesquisa do Estado de São Paulo (Fapesp – Processo no01/01661-3). Received on 25/5/2001. Reviewed on 22/10/2001. Approved on 14/11/2001.


INTRODUCTION
Giardia lamblia is a common human intestinal parasite of worldwide distribution found in both developed and developing countries.In endemic areas, high prevalence rates for G. lamblia infection have been reported in children, mainly among nursery and primary school children. 2,5,6,9,13ardia infection is usually diagnosed under light microscopy to identify either trophozoites or cysts in feces samples.In general, the examination of Giardia is performed in a single stool sample.However, as this parasite presents a variable pattern of excretion, misdiagnoses have been common and the actual prevalence may be underestimated.Microscopy examination of duodenal aspirate and jejunal biopsies is sometimes necessary.Since these two methods are invasive, they are rarely employed in children.
9][20] Most of these studies however were carried out among symptomatic giardiasis cases, but a large proportion of infected individuals are asymptomatic.
The present study was performed to compare serum anti-Giardia IgG antibody levels detected using indirect immunofluorescence (IIF) and ELISA to the frequency of infection in three stool examinations among children of day care centers in an endemic area.Also, the purpose was to verify whether serodiagnosis could be used to determine infection frequency in certain population groups, mainly asymptomatic cases.It is important to emphasize that immunologic assays were conducted using antigens prepared from trophozoites of an axenic strain isolated in Brazil and blood samples were collected on filter paper.

Stool and blood samples
The studied population consisted of 147 children ranging from 0 to 6 years old who attended three public day care centers in the city of Botucatu, state of São Paulo, Brazil.*Blood and feces samples were collected from children only after obtaining a formal consent signed by their parents.
Three stool specimens collected with a 7-day interval were examined using zinc sulfate flotation 3 and spontaneous sedimentation 14 methods.Blood samples obtained by finger tip lancet pricking were absorbed on Whatman filter paper #1, air-dried at room temperature and stored at -20°C until tested.A calibrated punch was applied to ensure discs measuring 1.32 cm, 2 which were incubated in 0.4 ml phosphate buffered saline solution (PBS) at pH 7.2 for 30 min.at room temperature, corresponding to a 1:20 serum dilution according to Souza & Camargo 17 (1966).Positive and negative controls were run in all assays.

Antigen preparations
Trophozoites of BTU-1 strain were employed to prepare antigens for IIF, ELISA, and absorption techniques.This strain was obtained from cysts found in the feces of a 7-year-old asymptomatic child and processed at the university' giardiasis laboratory (IB/ UNESP).Trophozoites were grown axenically in TYI-S-33 medium supplemented with bile 12 at 37°C.

Indirect immunofluorescence (IIF) assay
An equal volume of 1% formalin was added to trophozoites before overnight incubation at room temperature.The solution was washed three times with PBS, pH 7.2 and a concentration of about 40 organisms per 40x microscope field was used as antigen on the slide.Slides let to dry out at room temperature and stored at -70°C until used.
IIF assay was performed according to Vivesvara et al 21 with modifications.Eluates were used in a 1:20 serum dilution.One drop of each sample per well was added to antigen slides for 30 min.at 37°C.The slides were washed with PBS, pH 7.2 and goat anti-human IgG fluorescein isothiocyanate conjugates were added to each well at a 1:40 serum dilution.After 30 minutes, the slides were washed out and examined after mounting in glycerol buffered saline.A titer equals to or greater than 1:40 was considered positive.

ELISA assay
It was performed a modified assay according to Miotti et al 15 (1985).Trophozoites were centrifuged at 250 xg for 10 min.at 4°C, and the supernatant was discharged.The mass culture was re-suspended in PBS, pH 7.2 and washed twice as described above.The washed organisms were adjusted to a concentration of 10 4 trophozoites/50 µl, and then stored at -70°C until used.
Fifty microliters (50 µl) of trophozoite suspension (10 4 ) per well were used to coat plates.The plates *This study was approved by the Ethics Committee of the "Faculdade de Medicina da Universidade Estadual Paulista" and of the "Secretaria Municipal de Educação" were incubated overnight at 4°C, washed 3 times in PBS, pH 7.2 containing 0.05% Tween 20 (PBST), and then blocked for 1 hour with 150 µl/well of PBS with 5% skim milk.After 3 washes with PBST, 50 µl of each blood spot eluate 1:40 diluted were added to a well and incubated for 1 hour at room temperature.After washing, 50 µl of anti-IgG peroxidase conjugated diluted 1:4000 in PBS -1% skim milk were added to each well for 1 hour at room temperature.The enzymatic reaction was developed with 2 mg of o-phenylenediamine in 10 ml of phosphate citrate buffer, pH 5.0 upon addition of hydrogen peroxide.The reaction was interrupted by the addition of 25 µl of 4N sulfuric acid and the intensity of color was recorded as optical density (OD) at 492 nm wavelength.The cut-off value was calculated as the arithmetic mean of the negative control samples plus two standard deviations (X + 2SD).

Absorption technique
Antigens of Ascaris lumbricoides and Hymenolepis nana were prepared for absorption technique.A crude saline extract of homogenized worm was employed for obtaining A. lumbricoides antigens.H. nana antigen was obtained from a crude saline extract of worms collected from egg-inoculated mice.
The absorption technique was processed according to Wittner et al (1983). 20One-half milliliters of positive-titer eluate was incubated with 0.5 ml of sedimented antigen for 2 hours at 37°C and overnight at 4°C.After centrifugation at 600 xg for 20 minutes, the absorbed eluate was ready to use in both IIF and ELISA assays.

Data analysis
The correlation between microscopic examination of stool samples and serologic methods (IFI and ELISA) was analyzed using the McNemar test. 4The concordance between each serologic method (IIF and ELISA) and stool examination was determined considering both positive and negative cases for each method.
ELISA and IFI titers proportions according to stool examination results were analyzed using the Goodman test. 7,8I and ELISA sensitivity and specificity were determined regarding the microscopic visual interpretation of slides and OD values, respectively, where: Sensitivity % = Total samples both positive on microscopic examination and serologic test _________________________________________________________x 100 Total samples positive on microscopic examination Specificity % = Total samples both negative on microscopic examination and serologic test _________________________________________________________x 100 Total samples negative on microscopic examination

RESULTS
Microscopic examination of stool samples obtained from 147 children revealed that 93 (63.3%) were infected with G. lamblia.Only Giardia cysts were identif ied in all positive samples.No trophozoites were found in the stool samples.The highest percentages of infection were observed in children aged 12 to 47 months, and within this group the higher percentage of infected (18.3%) was seen in children aged 36 to 47 months.Other intestinal parasites found included Ascaris lumbricoides, Trichuris trichiura, Enterobius vermicularis, Ancylostomatidae, Strongyloides stercoralis, Hymenolepis nana, Entamoeba coli, and Blastocystis hominis.Forty (43%) of 93 stool samples with Giardia cysts and 20 (37%) of 54 negative samples had at least another intestinal parasite.
In IIF assay, antibody titers were invariably low ranging from 1:40 to 1:80 (Table 2).Of 93 positive samples in IIF, 64 were positive at 1:40 (Table 2).The analysis of IIF titer proportions according to stool examination revealed that the proportion of infected and seropositive children was higher at 1:40 dilution.For negative cases, the titers were lower than 1:40.
Most of the eluates obtained from children with positive stools were tightly bunched.There was an overlapping of these values with those of children without cysts in feces.The highest titers were observed among eluates obtained from children with cysts in feces.Some positive samples showed titers lower than the negative ones (Table 3).As for the analysis of ELISA titer proportions, there were no significant differences among <1:40, 1:40, 1:80 and 1:160 titers.In negative cases, no significant difference was observed between <1:40 and 1:40 titers, but there was a difference when they were compared to 1:80, 1:160, 1:320 and 1:640 titers.
The concordance between IIF assay and the 7-dayinterval examination of three stool samples was 76.9%.When ELISA was compared to microscopy examination the concordance was 59.9%.
All absorption tests performed with Giardia trophozoites against a positive eluate revealed a significant reduction in titers from the pre-absorption values.No variation was seen when the same eluate was absorbed against A. lumbricoides and H. nana antigens.

DISCUSSION
Though high prevalence rates of G. lamblia infection have been found in children of endemic areas, there are few serologic studies in these groups.In view of that Giardia is endemic in Brazil, and there are no seroepidemiological studies in children, the purpose of this study was to compare IIF and ELISA IgG of filter paper blood samples to three stool examinations with a 7-day interval between collection.

Most serologic studies have been able to detect
Giardia infection among symptomatic individuals.In general, the anti-parasite antibody response of symptomatic cases is stronger than that of asymptomatic individuals, and is related to the magnitude of cyst output. 16However, it is noteworthy that in endemic areas most infected individuals are asymptomatic carriers.
In the present study, when only asymptomatic individuals were enrolled, IIF showed a higher concordance than ELISA.IIF was more sensitive and showed higher specificity than ELISA.By its turn ELISA showed a high percentage of false-positive among samples of children without cysts in feces.Wittner et al 20 compared the same immunologic methods employing serum of symptomatic and asymptomatic individuals.It was observed that ELISA was also less specific than IIF, yielding more false-positive results among the groups without Giardia (58%).These authors raised the question whether this result could represent a measure of cross-reactivity or previous exposure to Giardia.
According to Soliman et al 16 (1998), in areas of high endemicity, the continuous, high level exposure of humans to the parasite early in life induces a stronger circulating anti-parasite antibody response that transects different age groups to the population.
Consistent with previous studies in endemic areas, 6,10 it was not possible to establish a correlation between infection and the presence of circulating antibodies.Besides, in cases with and without Giardia cysts in the feces, an overlap of titers was observed when both serologic methods were used to test the eluates.
Storing blood samples stored as dried spots on filter paper was tested by Al-Tukhi et al 1 (1993) and it was showed that this method is applicable to seroepidemiological studies in giardiasis.In this study, the main advantage of this method was to ease blood samples collection from 0 to 6-year old children, and facilitate their transport and storage.
As for the study results, IIF and ELISA did not show a satisfactory concordance with stool examination, but the prevalence of seropositivity was high for both methods, 62.8 % and 68.2 %, respectively, in comparison to stool examination.Gilman et al 6 (1985) in a study conducted in Bangladesh, a highly endemic Giardia area, observed a poor concordance between stool examination and serum antibody titers, showing that in areas with high infection rates, a single serum antibody titer cannot be used as a marker for Giardia infection.Thus it was suggested that antibody titers may be an useful epidemiological tool to determine the frequency of infection in endemic focus.According to Jokipii et al 11 (1988), despite exhaustive serologic studies, the results have been discrepant, varying from distinctive to overlapping titers in individuals with and without Giardia infection.It is important to note that not all discrepant results should be rejected since they can be due to: (1) the use of different strains of Giardia in antigen preparation; (2) enrolling only symptomatic patients in most of the studies; (3) choice of different serologic techniques and their various modifications.

Table 1 -
Comparison of microscopical examination of three stool samples, IIF and ELISA for Giardia lamblia infection.
*Values considered for calculation of concordance between each immunologic test and examination of stool samples.The concordance was determined with 95 per cent confidence interval.

Table 2 -
Indirect immunofluorescence titers for anti-Giardia IgG performed on filter paper blood samples from day-care children according to microscopical examination of three stool samples.

Table 3 -
ELISA titers for anti-Giardia IgG performed on filter paper blood samples from day-care children according to microscopical examination of three stool samples.The eluates were considered nonreactive if the OD< mean OD+2.0 SD of controls at a 1:40 dilution.