Assessing the validity of an ELISA test for the serological diagnosis of human fascioliasis in different epidemiological situations

Authors


Corresponding author M. Adela Valero, Departamento de Parasitología, Facultad de Farmacia, Universidad de Valencia, Av. Vicente Andrés Estellés s/n, 46100 Burjassot, Valencia, Spain. Tel: +34 963544490; Fax: +34 963544769; E-mail: madela.valero@uv.es

Abstract

Objectives  To improve the diagnosis of human fascioliasis caused by Fasciola hepatica and Fasciola gigantica, we evaluated the diagnostic accuracy of an enzyme-linked immunosorbent assay (ELISA), with Fasciola antigen from the adult liver fluke, for the detection of IgG against fascioliasis in human sera.

Methods  The sera of 54 fascioliasis cases, originating from three endemic areas, were used in this evaluation: (i) a hyperendemic F. hepatica area where humans usually shed a great number of parasite eggs in faeces (11 sera); (ii) an epidemic F. hepatica area where humans usually shed small amounts of parasite eggs (24 sera) and (iii) an overlap area of both Fasciola species and where human shedding of parasite eggs in faeces is usually scarce or non-existent (19 sera). One hundred and sixty-eight patients with other parasitic infections and 89 healthy controls were also analysed.

Results  The respective sensitivity and specificity of this assay were 95.3% (95% confidence intervals, 82.9–99.2%) and 95.7% (95% confidence intervals, 92.3–97.5%). No correlation between egg output and the OD450 values of the F. hepatica IgG ELISA test was observed.

Conclusions  This test could be used both as an individual serodiagnostic test for human fascioliasis when backed up by a compatible clinical history together with a second diagnostic technique for other cross-reactive helminth infections, and in large-scale epidemiological studies of human fascioliasis worldwide.

Abstract

Objectifs:  Pour améliorer le diagnostic de la fasciolose humaine causée par Fasciola hepatica et Fasciola gigantica, nous avons évalué la précision diagnostique d’un test immunoenzymatique (ELISA), avec l’antigène de Fasciola de la douve adulte du foie, pour la détection des anticorps IgG contre la fasciolose dans des sera humains.

Méthodes:  Les sera de 54 cas de fasciolose, provenant de trois régions endémiques, ont été utilisés dans cette évaluation: (i) une région hyper endémique pour F. hepatica où les gens excrètent généralement un grand nombre d’œufs du parasite par les selles (11 sera), (ii) une région épidémique pour F. hepatica où les gens excrètent généralement de petites quantités d’œufs du parasite (24 sera) et (iii) une région de chevauchement entre les deux espèces de Fasciola, où l’excrétion des œufs de parasites de l’homme dans les selles est généralement rare, voire inexistante (19 sera). 168 patients souffrant d’autres infections parasitaires et 89 témoins sains ont également été analysés.

Résultats:  La sensibilité et la spécificité respectives de ce test étaient de 95,3% (IC95%: 82,9 à 99,2%) et 95,7% (IC95%: 92,3 à 97,5%). Aucune corrélation n’a été observée entre l’excrétion d’œufs et les valeurs de DO450 du test ELISA basé sur les IgG de F. hepatica.

Conclusions:  Ce test pourrait être utiliséà la fois comme un test de sérodiagnostic des individus pour la fasciolose humaine lorsqu’il est soutenu par une histoire clinique compatible, comme une seconde technique de diagnostic pour d’autres helminthiases à réaction croisée et utiliséà grande échelle pour des études épidémiologiques de la fasciolose humaine à travers le monde.

Abstract

Objetivos:  Con el fin de mejorar el diagnóstico de la fascioliasis humana causada por la Fasciola hepatica y la Fasciola gigantica, hemos evaluado la precisión diagnóstica de una prueba de ELISA, con el antígeno Fasciola, de la duela hepática adulta, para la detección de IgGs en suero humano.

Método:  Se utilizó el suero de 54 casos de fascioliasis, originarios de tres áreas: (i) un área hiperendémica para F. hepatica en donde los humanos usualmente eliminan un gran número de huevos del parásito en las heces (11 sueros); (ii) un área epidémica para F. hepatica, en donde los humanos usualmente eliminan pequeñas cantidades de huevos de parásito (24 sueros); y (iii) un área en donde se sobreponen ambas especies de Fasciola y en donde la eliminación de huevos de parásito en heces humanas es usualmente escasa o no se da (19 sueros). También se analizaron sueros de 168 pacientes con otras infecciones parasitarias y 89 controles sanos.

Resultados:  La sensibilidad y especificidad del ensayo fue respectivamente del 95.3% (IC 95%, 82.9-99.2%) y del 95.7% (IC 95%, 92.3-97.5%). No se observó ninguna correlación entre el resultado de huevos y los OD450 en la prueba de ELISA para IgG contra F.hepatica.

Conclusiones:  Esta prueba podría utilizarse tanto como una prueba serodiagnóstica individual para la fascioliasis humana cuando esté respaldada por una historia clínica compatible, y una segunda técnica de diagnóstico para otras infecciones helmínticas de reacción cruzada, y en estudios epidemiológicos a gran escala sobre fascioliasis humana en todo el mundo.

Introduction

Fascioliasis is a parasitic disease caused by liver fluke species of the genus Fasciola: Fasciola hepatica and Fasciola gigantica. While F. hepatica is found on a global scale, F. gigantica can be found in humans and animals in tropical regions of Africa and Asia (Mas-Coma et al. 2009a). However, in some parts of Africa and Asia, the two species overlap (Ashrafi et al. 2006; Periago et al. 2008; Mas-Coma et al. 2009a). Today, fascioliasis is considered an important human disease and several geographic areas have been described as endemic for the disease in humans, including hypoendemic (prevalence < 1%), mesoendemic (prevalence between 1% and 10%) and hyperendemic (> 10%) situations, with intensities ranging from low to very high, and estimates of up to 17 million people infected worldwide. This may even be an underestimate if the total lack of data from numerous Asian and African countries is considered (Mas-Coma et al. 1999).

Moreover, in the last two decades, human fascioliasis has been emerging in many regions, a phenomenon that has partly been related to climate change (Mas-Coma et al. 2009b). Recent studies have shown that the high pathogenicity of this disease is not only restricted to the acute phase, but also to long-term liver fluke infection (Valero et al. 2000, 2003, 2006, 2008), the immune-modulation of fasciolids in the acute phase (Brady et al. 1999), and their immune suppression effect in chronic and advanced chronic phases (Gironès et al. 2007). All this appears to be in the background of usual co-infections with other parasitic and infectious diseases (Mas-Coma et al. 2005). Considering this scenario of increasing concern, WHO decided to launch a worldwide initiative against this disease (WHO 2007, 2008). Therefore, the capacity to diagnose human fascioliasis correctly becomes imperative.

The present study assesses a qualitative microtiter strip-based enzyme-linked immunosorbent assay (ELISA) for the detection of IgG class antibodies against F. hepatica in human serum, a solid phase enzyme immunoassay based on the sandwich principle, known as ‘DRG Fasciola hepatica IgG (human) ELISA’, in the following situations: (i) a hyperendemic F. hepatica area where humans usually shed large amounts of parasite eggs in faeces, (ii) an epidemic F. hepatica area where humans usually shed small amounts of parasite eggs and (iii) an overlap area of both Fasciola species where human shedding of parasite eggs in faeces is usually scarce or non-existent. Fascioliasis positivity of the samples used had been verified through egg detection (Kato–Katz technique) or the ELISA Cathepsin L1 (CL1) protease test. ELISA CL1 is used here as the serological gold standard method in samples from areas where humans usually shed small amounts of parasite eggs or where parasite eggs in faeces are usually scarce or non-existent.

ELISA CL1 was previously successfully employed for the diagnosis of human fascioliasis in the northern Bolivian Altiplano (O’Neill et al. 1998) and in Iran (Rokni et al. 2002), showing it to be highly specific (100%) and sensitive (100%).

Material and methods

Clinical samples

Sera were obtained from the serum bank of the Department of Parasitology, Faculty of Pharmacy, University of Valencia, Spain (2004–2008). Each serum was aliquoted and stored at −80 °C until used. The sera were classified in three groups: (i) 54 positive fascioliasis sera through ELISA CL1 and eggs in faeces quantified through the Kato–Katz technique, originating from three different epidemiological situations; (ii) 168 serum samples from individuals with parasitic/fungal diseases other than fascioliasis (to evaluate potential cross-reactivity) were obtained from clinically, serologically and sometimes parasitologically confirmed cases and (iii) 89 sera from parasitologically negative cases were used from serum collections stored at Centro Nacional de Microbiología (ISCIII, Madrid, Spain) and Servicio de Microbiologia, Hospital Universitario y Politecnico ‘La Fe’ (Valencia, Spain). The age and gender characteristics of the individuals are detailed in Table 1.

Table 1.   Gender and age characteristics of individuals whose serum samples were used in the validation
 TotalMaleFemaleInd.
N Average SD (min–max)NAge
Average SD (min–max)
N Age
Average SD (min–max)
N Age
Average SD (min–max)
  1. *Amoebiasis 12; leishmaniasis 3; malaria 12; toxoplasmosis 10; schistosomiasis 15; hydatidosis 15; hymenolepiasis 1; taeniasis/cysticercosis 15; trichinellosis 15; strongyloidiasis 18; ascariasis 1; toxocariasis 8; anisakiasis 25; loasis 2; gnathostomiasis 10; trichuriasis and ascariasis coinfection 1; malaria, trichuriasis and ascariasis coinfection 2; Entamoeba coli, trichuriasis, ascariasis and ancylostomiasis coinfection 1; trichuriasis, ancylostomiasis and enterobiasis coinfection 1; histoplasmosis 1.

Individuals positive fascioliasis 54(36.30 ± 1.09)
(7–75)
25(34.25 ± 1.21)
(7–74)
29(38.00 ± 1.00)
(10–75)
Individuals with parasitic/fungal diseases other than fascioliasis168*(38.15 ± 1.07)
(1–89)
82(40.42 ± 1.11)
(1–89)
83(36.05 ± 1.03)
(1–89)
331.50 ± 0.67
(23–40)
Individuals parasitologically negative 89(40.28 ± 0.81)
(1–73)
55(42.68 ± 0.68)
(22–73)
33(36.33 ± 0.95)
(1–73)
1Unknown
Unknown
Total        

ELISA Cathepsin L1

Cathepsin L1 (CL1) was obtained from the Molecular Parasitology Laboratory, School of Biotechnology, Faculty of Health and Science, Dublin City University, Dublin, Ireland. Application of the ELISA CL1 test was carried out as previously described (O’Neill et al. 1999). Briefly, 100 μl of CL1 antigen (5 μg/ml) were incubated overnight at 37 °C in wells of microtiter plates (Nuclon, Kamstrup, Roskilde, Denmark). The wells were blocked with a solution of 2% bovine serum albumin/0.1% Tween 20 in phosphate-buffered saline for 30 min at 37 °C, and human sera (1:100 dilution) were then added into duplicate wells. Bound human antibodies were detected using biotin-conjugated anti-human IgG4 (1:2,000 dilution), avidin-conjugated alkaline phosphatase, and the substrate 2,2-azino-bis(3-ethylbenzthiazoline-6-suflonic acid) (all from Sigma Chemical Co., Poole, Dorset, UK). Plates were read after 20 min on a multiscan ELISA plate reader at an absorbance of 405 nm (Bio-Rad, Model 680, Hercules, CA, USA). The cut-off point between the clusters was set at 3.09 standard deviations from the mean of the control sera absorbance and was 0.098.

DRG kit

DRG Instruments GmbH, Germany, provided all commercial kits used in this study. The test is based on Excretion/Secretion antigens of F. hepatica predominantly containing fluke cysteine proteases (Salimi-Bejestani et al. 2005). All steps were carried out according to the manufacturer’s instructions. Human sera, negative and positive controls and cut-off control (CO) (1:100 dilution) were added into triplicate wells. Plates were read after 15 minutes on a multiscan ELISA plate reader at an absorbance of 450 nm (Bio-Rad, Model 680, Hercules, CA, USA). The test run may be considered valid if substrate blank has an absorbance value below 0.100, negative control below 0.200, CO control between 0.250 and 0.750 and positive control above 0.600. The results are given as the mean of the optical density (OD) obtained from triplicate samples expressed as a percentage of the CO, using the following formula:

image

A serum is considered positive when its absorbance value is above 10% of CO.

The results in DRG Units (DU) were calculated according to the following formula:

image

The results were negative if DU < 9, and positive if DU > 11.

To avoid any bias, all sera when being processed were blinded with regard to the adscription of the aforementioned three groups (fascioliasis positive, other parasitoses or negative sera).

Ethical aspects

The present study was approved by University of Valencia Ethics Committee.

Data analysis

The diagnostic sensitivity and specificity values were calculated (Fletcher & Fletcher 2005) with its 95% confidence interval (95% CI; EPIINFO). The positive predictive value (PPV) was calculated using the following formula:

image

The negative predictive value (NPV) was calculated using the following formula:

image

Student’s t-test was calculated using SPSS software.

Results

The mean of DU values for all serum groups is shown in Figure 1A,B. DU values were significantly higher in the proven fascioliasis group than in the groups of parasitic diseases other than fascioliasis and the healthy controls, respectively (< 0.001).

Figure 1.

 Mean DRG Units (DU) obtained with the DRG F. hepatica IgG ELISA test in fascioliasis patients from different epidemiological situations. (A) Endemic F. hepatica area where humans usually shed small amounts of parasite eggs; (B) an overlap area of both Fasciola species where human shedding of parasite eggs in faeces is usually scarce or non-existent. Data points are the mean optical density (OD) obtained from triplicate samples expressed as a percentage of the cut-off (CO), using the following formula: inline image. A serum is considered positive when its absorbance value is above 10% of CO. The results in DRG Units (DU) were calculated according to the following formula: inline image. The results were negative if DU < 9, and positive if DU > 11 (dotted line).

The numbers of true positive and false negative test results in fascioliasis cases of humans with and without detection of eggs in their faecal samples are provided in Table 2.

Table 2.   Diagnostic characteristics of the F. hepatica IgG ELISA test applied to fascioliasis patients’ sera from different disease endemic areas, sera from patients with helminthiases other than fascioliasis and negative sera from healthy subjects
 True positiveFalse negativeTrue negativeFalse positive
  1. *Gold standard determined according to the Kato–Katz technique.

  2. †Gold standard determined according to the ELISA CL1 test.

Hyperendemic F. hepatica area where humans usually shed a great deal of parasite eggs in faeces*110
Epidemic F. hepatica area where humans usually shed small amounts of parasite eggs†222
Overlap area of both Fasciola species where human shedding of parasite eggs in faeces is usually scarce or non-existent†190
Total sera with fascioliasis522
Individuals with parasitic/fungal diseases other than fascioliasis15810
Individuals parasitologically negative881
Total sera without fascioliasis24611

In the samples from a hyperendemic F. hepatica area egg detection in faeces through the Kato-Katz technique was used as reference standard, and the sensitivity of the F. hepatica IgG ELISA test was 100% (67.9–100%). No correlation between egg output − a measure of infection intensity − and the OD450 values of the F. hepatica IgG ELISA test was observed in this study area (Figure 2).

Figure 2.

 DRG F. hepatica IgG ELISA test and intensity of F. hepatica infection. Data points represent the mean absorbance at 450 nm from copropositive patients from a hyperendemic F. hepatica area where humans usually shed a great deal of parasite eggs in faeces. epg = egg count per gram of faeces.

In the samples from an epidemic F. hepatica area CL1 ELISA was used as reference standard. Sensitivity was 91.7% (95% confidence intervals, 71.5–98.5%), detecting 2 false negative cases of which one was egg positive in faeces, while the other was not.

In the samples from an overlap area of both Fasciola species CL1 ELISA was also used as reference standard. Sensitivity was 100% (95.9–100%). Populations of the two latter endemic areas were grouped as result of the ELISA CL1 and the F. hepatica IgG ELISA test evaluation as follows: fascioliasis positive cases by CL1 (Fas+)/seropositive by F. hepatica IgG ELISA test (test+) and fascioliasis positive cases by CL1 (Fas+)/seronegative by F. hepatica IgG ELISA test (test-) (Figure 1A,B). Statistical differences between the results for sensitivity values from the two latter situations were not found, and true positive cases (41) and false negative cases (2) for these two areas together were used for the calculation of sensitivity, presenting a value of 95.3% (82.9–99.2%).

Specificity was calculated from the results in Table 2 providing the numbers of false negative and true positive test results in infected sera other than fascioliasis and negative cases. These results can be grouped as follows: infected positive cases other than fascioliasis (infected+)/seropositive by F. hepatica IgG ELISA test (test+) and infected positive cases other than fascioliasis (infected+)/seronegative by F. hepatica IgG ELISA test (test-) (Figure 3A). Table 2 also shows the total number of sera without fascioliasis, and these values were used to calculate specificity, presenting a value of 95.7% (92.3–97.5%). Ten serum samples from a heterologous infection cross-reacted with the kit, mainly originating from patients with helminthiases, namely schistosomiasis (4 samples), hydatidosis (2 samples), taeniasis/cysticercosis (1 sample), trichinosis (1 sample), strongyloidiasis (1 sample), and histoplasmosis (1 sample) (see Figure 3B). Theoretical PPVs and NPVs vs fascioliasis prevalence are represented in Figure 4, showing the expected PPVs and NPVs depending on whether the test was used in low, medium or high prevalence scenarios.

Figure 3.

 Mean DRG Units (DU) obtained with the DRG F. hepatica IgG ELISA test in (A) patients with parasitoses other than fascioliasis (B) and negative control sera. Data points represent the mean absorbance at 450 nm obtained from three replicates of each serum tested. The dotted line represents the cut-off value of OD at 450 nm. Data points are the mean optical density (OD) obtained from triplicate samples expressed as a percentage of the cut-off (CO), using the following formula: inline image. A serum is considered positive when its absorbance value is above 10% of CO. The results in DRG Units (DU) were calculated according to the following formula: inline image. The results were negative if DU <9, and positive if DU > 11 (dotted line).

Figure 4.

 Theoretical PPVs and NPV values vs fascioliasis prevalence, showing the expected PPVs and NPV values if the test were used in low (below 1%), medium (between 1% and 10%) or high (above 10%) prevalence scenarios (expressed in vertical lines).

Discussion

Coprological analysis, based on the identification of eggs found in stools, duodenal contents or bile analysis is still commonly employed to diagnose human fascioliasis (Valero et al. 2009a), despite the overwhelming consensus that this method is not wholly reliable (Hillyer 1999) for several reasons. Eggs are not detected until the latent period of infection when much of the liver damage has already occurred, and similarly patients may not shed eggs in faeces in ectopic cases known in both infection by F. hepatica and F. gigantica (Le et al. 2007). Additionally, eggs are released sporadically from the bile ducts and hence stool samples of infected patients may not necessarily contain eggs (Mas-Coma et al. 1999).

However, coprological methods of egg detection are still usually considered the gold standard, although unfortunately they are not applicable in endemic areas where: (i) humans do not shed eggs in faeces, (ii) the shedding is often very low and difficult to detect and also (iii) in cases of communities that do not supply stool samples due to ethnic/cultural customs. Hence, it is essential to have serological techniques also available when diagnosing the disease. Therefore, in this study, besides Kato–Katz, another serological test, the ELISA CL1 test, was also used as gold standard.

Serodiagnosis of fascioliasis in human and animal species has been successfully carried out employing several antigenic fractions of Fasciola (Mezo et al. 2003; Sánchez-Andrade et al. 2008; Demerdash et al. 2011), purified antigens (O’Neill et al. 1998; Rokni et al. 2002), and recombinant antigens (O’Neill et al. 1999; Carnevale et al. 2001). Cathepsins L are the most frequently used target antigens for detecting anti-Fasciola antibodies (Carnevale et al. 2001; Rokni et al. 2002; Mezo et al. 2004, 2007, 2010; Intapan et al. 2005; Wongkham et al. 2005; Valero et al. 2009b; Muiño et al. 2011), as circulating antibodies to these molecules remain at high levels for long periods (Valero et al. 2009b).

The commercial DRG test was evaluated in cattle, obtaining a sensitivity and specificity of 98% (96–100%) and 96% (93–98%) respectively at a cut-off value of 15% positivity (Salimi-Bejestani et al. 2005). The sensitivity and specificity values of the DRG test herein agree with the results obtained by other authors for F. hepatica IgG ELISA in-house assays: 100% and 100% (O’Neill et al. 1998; Rokni et al. 2002), 92.4%, and 83.6% (Espinoza et al. 2007), 97.2% and 100% (Rahimi et al. 2011), 97.0% and 96.6% (Cornejo et al. 2010) and 100 and 95.6% (Figueroa-Santiago et al. 2011). Commercial kits offer advantages over in-house assays: they save time and provide quality control reagents for better reproducibility within and between laboratories, which hampers the comparison between in-house and commercial tests. Furthermore, only very few commercial kits, such as the DRG F. hepatica IgG ELISA test evaluated here, are presently available for the diagnosis of human fascioliasis. Comparing DRG test results only with commercial assays results in sensitivity and specificity below the 100% sensitivity, respectively, specificity obtained by SeroFluke strips (Martínez-Sernández et al. 2011).

The results obtained show the test to be highly sensitive and specific. The sensitivity of this assay did not detect differences in positive samples from different epidemiological situations. Nevertheless, PPVs calculated for diverse epidemiological situations are very different. Thus, PPVs in hypoendemic areas were below 2.2%, in mesoendemic situations PPVs oscillated between 2.2% and 71.2%, while in hyperendemic situations PPVs were above 71.17%, making this test recommendable in such situations. Contrarily, NPVs calculated for diverse epidemiological situations are similar. NPVs for hypoendemic areas reached values above 99.9%, in mesoendemic situations NPVs oscillated between 99.5% and 99.9%, while in hyperendemic situations NPVs were below 99.5%.

In conclusion, the DRG test could be used both as (i) an individual serodiagnostic test for human fascioliasis when backed up by a clinical history, together with a second diagnostic technique in certain situations of possible cross-reactions (patients co-infected with other helminths) and (ii) in future large-scale epidemiological studies of human fascioliasis worldwide (recommended for mass screening, especially when considering its convenient handling).

Acknowledgments

CL1 was supplied by Molecular Parasitology Laboratory, School of Biotechnology, Faculty of Health and Science, Dublin City University, Dublin, Ireland. I. Perez-Crespo received a fellowship of the Programa de Ayudas Predoctorales de Formación de Personal Investigador, Spanish Ministry of Education and Science. Financial support was obtained from the Spanish Ministry of Science and Innovation, Madrid, Spain; the Red de Investigación de Centros de Enfermedades Tropicales, the Ministry of Health, Madrid, Spain; and the Conselleria de Empresa, Universidad y Ciencia, Valencia, Spain. The authors also acknowledge the English stylistic help by Mr. R. Wilk (Valencia, Spain).

Ancillary