Introduction

Diarrhea is one of the main causes of morbidity and mortality in the world in particular in children in developing countries (WHO 2005). The etiological agents of diarrhea include viruses, bacteria, and parasites (Thielman and Guerrant 2004). Entamoeba histolytica, Giardia intestinalis, and Cryptosporidium species are the most common diarrhea-causing protozoan parasites, which induce at least in part indistinguishable clinical presentations (Walsh 1986). The WHO estimates that ~50 million people worldwide suffer from invasive amebic infection each year, resulting in 40,000–100,000 deaths annually (WHO 1997; Diamond and Clark 1993; Petri et al. 2000). G. intestinalis is the most common protozoan infection of the intestinal tract worldwide. Each year, 500,000 new cases are reported and about 200 million people develop symptomatic giardiasis (Minvielle et al. 2004). Cryptosporidiosis is a frequent cause of diarrheal disease in humans, in particular in immunocompromised patients. Moreover, Cryptosporidium infections occur in children younger than 5 years of age, with a peak in children younger than 2 years of age (Tumwine et al. 2003; Steinberg et al. 2004).

In Egypt, the three parasites are widely prevalent. Prevalence rates ranging from 16.2 to 57.1 % were recorded for E. histolytica/dispar (El-Kadi et al. 2006; Abd-Alla and Ravdin 2002; Abdel-Hafeez et al. 2012). Rates varying between 10.0 and 34.6 % were recorded for G. intestinalis (El Naggar et al. 2006; Foronda et al. 2008; Baiomy et al. 2010; Abdel-Hafeez et al. 2012). For Cryptosporidium species, several studies reported prevalence rates ranging from less than 5.0–31.1 % (Ibrahim et al. 1997; Abdel-Messih et al. 2005; Mousa et al. 2010; Abdel Kader et al. 2012).

Knowledge of the etiology of diarrhea is important for epidemiological surveillance and for correct treatment. Traditionally, the three parasites have been identified by simple microscopic methods. As expertise in stool microscopy is waning and multiple sampling, species-specific concentration and staining methods are needed to improve its performance, many of the infections were missed and some are overestimated due to similarities in morphologies as in the case of the E. histolytica/dispar/moshkovskii complex (Hamzah et al. 2010). Alternative approaches have been developed to improve the diagnosis of enteric parasitic diseases, including visualization by fluorescent-labeled antibodies and copro-antigen-detection assays, but many of these tests still lack sensitivity and specificity (Murray and Capello 2008; Ndao 2009). Meanwhile, polymerase chain reaction (PCR) methods for detecting intestinal parasites are increasingly available and exhibit excellent sensitivity and specificity compared to conventional methods such as microscopy and antigen detection assays (Webster et al. 1996; Sanuki et al. 1997; Haque et al. 1998; Fisher et al. 1998; Ghosh et al. 2000; Stark et al. 2008).

Real-time PCR (RT-PCR) is a very attractive technique for laboratory diagnosis of infectious diseases, as the methodology does not require post-PCR downstream analysis, leading to shorter turnaround times. Moreover, RT-PCR substantially reduces the risk of amplicon contamination of the laboratory and decreases the cost for reagents (Klein 2002). In addition, real-time PCR is a quantitative method that allows the determination of the parasite burden. Several real-time PCR assays have been developed to detect the common enteric protozoan parasites (Blessmann et al. 2002; Verweij et al. 2003; Qvarnstrom et al. 2005; Calderaro et al. 2010; Hadfield et al. 2011). A multiplex real-time PCR for the simultaneous detection of E. histolytica, G. intestinalis, and Cryptosporidium parvum in fecal samples was recently described (Verweij et al. 2004; Haque et al. 2007; ten Hove et al. 2007). The assay has been found to be quite sensitive and specific and is able to detect each parasite individually.

In this study, the multiplex real-time PCR, as a single test tube assay, was used for the detection of E. histolytica, G. intestinalis, and Cryptosporidium spp. among Egyptian patients complaining of diarrhea and in a representative control group. The results were compared with those obtained by routine microscopy.

Material and methods

Fecal specimens

A case–control study was conducted in which fresh stool samples were collected from 396 patients complaining of diarrhea and 202 apparently healthy individuals. The samples were collected over a period of 1 year from October 2010 to October 2011 in Cairo and the Egyptian governorates Fayoum and Benha, respectively. The patients and the healthy controls aged between 6 months and 60 years.

Microscopy

After collection, stool samples were divided into two portions; the first portion was preserved frozen at −20 °C for further processing by real-time PCR, the second portion was examined microscopically by direct saline and/or iodine mounts and after concentration by formol-ethyl acetate technique. Modified Ziehl–Neelsen was performed on direct fresh smears as well as on formol-ethyl acetate concentrates to detect Cryptosporidium oocysts.

DNA extraction

Of the stool sample, 0.2 g was used for extraction of DNA using the QIAamp DNA stool mini kit (QIAGEN, Hilden, Germany) according to the manufacturer’s instructions. E. histolytica control DNA was obtained from an axenic culture of the strain HM-1: IMSS. G. intestinalis DNA was isolated from purified cysts and C. parvum DNA was isolated from the purified oocysts. In each sample, 103 PFU of phocin herpesvirus 1 (PhHV-1) per milliliter was added to the isolation lysis buffer to serve as an internal control (Niesters 2002).

Primers and probes

The primers and probes described by Verweij et al. (2004) were used in the present study. They were all purchased from Eurofins MWG Operon, Germany. For E. histolytica and Entamoeba dispar, the primers Ehd-38F (5′-ATTGTCGTGGCATCCTAACTCA-3′) and Ehd-88R (5′-GCGGACGGCTCAT TATAACA-3′) targeting the small subunit of ribosomal RNA gene (SSU rRNA) of both E. histolytica and E. dispar such that to amplify a 172-bp fragment inside the gene. The Taqman probes; the minor groove binding (MGB) probe histolytica-96 T (JOE-5′-TCATTGAATGAATTGGCCATTT-3′-BHQ1) and the MGB probe dispar-96 T (Cy5-5′-TTACTTACATAAATTGGCCACTTTG-3′-BHQ2; Verweij et al. 2003), specifically detect the E. histolytica and E. dispar amplification products, respectively.

For G. intestinalis, the primers Giardia-80F (5′-GACGGCTCAGGA CAACGGTT-3′) and Giardia-127R (5′-TTGCCAGCGGTGT CCG-3′) targeting SSU rRNA such that to amplify a 62-bp fragment inside the gene. The Taqman probe double-labeled Giardi-105 T (Fam-5′-CCCGCGGCGGTCCCTGCTAG-3′-Tamra) specifically detects the amplification products.

For C. parvum, the primers designed by Fontaine and Guillot (2002) were used. The primers CrF (5′-CGCTTCTCTAGCCTTTCATGA-3′) and CrR (5′-CTTCACGTGTGTTTGCC AT-3′) targeting the genomic DNA sequence such that to amplify a 138-bp fragment inside the C. parvum-specific 452-bp fragment. The double-labeled probe Crypto (Rox-5′-CCAATCACAGAATCATCAGAATCGACTGGTATC-3′-BHQ2) specifically detects the amplification product.

As an internal control to detect possible PCR inhibition of amplification by stool contents, a specific primer and probe set, consisted of a forward primer PhHV-267s (5′-GGGCGAATCACAGATTGAATC-3′), a reverse primer PhHV-337as (5′-GCGGTTCCAAACGTACCAA-3′), and the specific double-labeled probe PhHV-305tq (Cy5.5-5′-TTTTTATGTGTCCGCCACCATCTGGATC-3′-BBQ), targeting PhHV-1 were included with each run.

PCR amplification and detection

Amplification reactions were performed in a volume of 25 μL with Qiagen HotstarTaq master mix, 5 mM MgCl2, 3.125 pmol of each E. histolytica/dispar-specific primers, 3.125 pmol of each G. intestinalis-specific primers, 12.5 pmol of each C. parvum-specific primers, 1 pmol of each PhHV-1-specific primers, 4.375 pmol of E. histolytica-specific MGB-TaqMan probe, 4.375 pmol of E. dispar-specific MGB-TaqMan probe, 0.25 pmol of G. intestinalis-specific double-labeled probe, 4.375 pmol of C. parvum-specific double-labeled probe, and 1.25 pmol of PhHV-1-specific double-labeled probe. Amplification consisted of 3 min at 95 °C followed by 40 cycles of 30 s at 95 °C, 30 s at 55 °C, and 30 s at 72 °C. Fluorescence was measured during the annealing step of each cycle. Amplification, detection, and data analysis were performed with the Rotor gene 6000 real-time detection system.

Results

Table 1 shows the comparison of direct microscopy versus formol ethyl-acetate concentration technique for detection of E. histolytica/dispar, G. intestinalis, and Cryptosporidium spp. among 396 diarrheal fecal samples along with 202 fecal samples from healthy controls. While no significant difference was found between the two techniques for the detection of G. intestinalis and Cryptosporidium spp., formol ethyl-acetate concentration exhibited higher significant difference for detection of E. histolytica/dispar infections.

Table 1 Comparison of direct smear versus formol ethyl-acetate concentration technique for the detection of enteric protozoa among 396 patients with diarrhea and 202 control subjects
Full size table

The results of examining diarrheal fecal samples along with fecal samples from healthy controls either by microscopy after formol ethyl-acetate concentration or multiplex real-time PCR for the simultaneous detection of E. histolytica, E. dispar, G. intestinalis, and Cryptosporidium spp. are shown in Table 2. Among diarrheal cases, microscopy detected G. intestinalis as mono infection in 110 (27.8 %) and as combined infection with E. histolytica/dispar in 11 cases (2.8 %). Sole E. histolytica/dispar and Cryptosporidium spp. infections were detected in 32 (8 %) and 4 (1 %) cases, respectively. Two hundred thirty-nine cases (60.4 %) were found negative for the three protozoan parasites suggesting other etiologies for diarrhea. In total, microscopy detected G. intestinalis in 30.5 % (121 out of 396), E. histolytica/dispar in 10.8 % (43 out of 396), and Cryptosporidium spp. in 1 % (4 out of 396) of the diarrheal cases. Among the control subjects, G. intestinalis was detected in only one case (0.25 %) whereas E. histolytica/dispar and Cryptosporidium spp. were not detected in any of the control subjects.

Table 2 Results of microscopy and multiplex real-time PCR for the detection of enteric protozoa among 396 patients with diarrhea and 202 control subjects
Full size table

Within the 396 fecal samples from patients with diarrhea, multiplex real-time PCR revealed mono-infections with G. intestinalis, E. dispar, and Cryptosporidium spp. in 141 (35.6 %), 5 (1.26 %), and 8 (2.0 %) of the cases, respectively. In addition, six combined infections were identified comprising two (0.5 %) cases with G. intestinalis and E. dispar, three (0.75 %) cases with G. intestinalis and Cryptosporidium spp., and one (0.25 %) case containing all three protozoan parasites. It is noteworthy that E. histolytica was not detected in any of the diarrheal patients. Within the 202 controls, G. intestinalis was present in four (2.0 %), E. histolytica in one (0.5 %), and Cryptosporidium spp. in three (1.5 %) of the individuals investigated. It is worth noting that the only case of E. histolytica detected by real-time PCR was among the control subjects, whereas all the eight Entamoeba cases detected among the diarrheal patients were of the nonpathogenic E. dispar.

The validity of microscopy compared to multiplex real-time PCR for the detection of the three protozoan parasites is shown in Table 3. Sensitivities varied from 33.3 % for Cryptosporidium spp. to 57.8 % for G. intestinalis and 100 % for E. dispar. Specificities varied from 85.5 % for G. intestinalis to 91 % for E. dispar, and 100 % for Cryptosporidium spp. Positive and negative predictive values varied from 18.6 to 100 % as shown in Table 3. Out of the 43 microscopy-positive E. histolytica/dispar samples, only eight were true E. dispar by real-time PCR. All of the 353 cases microscopically negative for E. histolytica/dispar were also negative by real-time PCR. Out of the 121 microscopy-positive G. intestinalis samples, only 85 were true positive by real-time PCR. Out of the 275 cases microscopically negative for G. intestinalis, 62 were positive by real-time PCR. Out of the 394 microscopically negative Cryptosporidium samples, eight were found to be true positive by real-time PCR. All the four microscopically positive Cryptosporidium samples were also positive by real-time PCR. Comparison of the prevalence of the different parasites between the three study sites within Egypt indicated no significant differences (Table 4).

Table 3 Sensitivity, specificity, positive and negative predictive values of microscopy compared to multiplex real-time PCR for detection of E. histolytica/dispar, G. intestinalis, and Cryptosporidium spp. in the stool of 396 diarrheal patients
Full size table
Table 4 Prevalence of E. histolytica/dispar, G. intestinalis, and Cryptosporidium spp. in diarrheal patients form three Egyptian Governorates
Full size table

Discussion

Diarrhea is an important cause of morbidity and mortality in the world. G. intestinalis, Cryptosporidium spp., and E. histolytica are the most common diarrhea-causing parasitic protozoa (Pierce and Kirkpatrick 2009). Microscopy is the most commonly used method for the routine diagnosis of these parasites in developing countries, however, microscopy lack sensitivity and specificity (Utzinger et al. 2010). Replacing microscopy with the more sensitive and specific nucleic acid-based methods is hampered by the cost in developing countries. Multiplexing the detection of more than one parasite in a single test by real-time PCR has been found to be very effective (Verweij et al. 2004; Haque et al. 2007; ten Hove et al. 2007; Amar et al. 2007; Stark et al. 2011; Bruijnesteijn van Coppenraet et al. 2009; Taniuchi et al. 2011). Adopting such approached in developing countries, at least in reference laboratories, would decrease the cost and ensure rapid and accurate diagnosis of diarrheal causing protozoa. In the present study, stool samples collected from 396 patients complaining of diarrhea and 202 apparently healthy controls were tested by microscopy in Egypt. Frozen portions of the same samples were tested by multiplex real-time PCR in the Department for Molecular Parasitology at The Bernhard Nocht Institute for Tropical Medicine, Hamburg, Germany, which represents the German National Reference Centre for tropical infections.

The results indicate that among diarrheal patients in Egypt G. intestinalis is the most common protozoan parasite, with prevalence rates of 30.5 and 37.1 %, depending of the method used (microscopy vs. multiplex real-time PCR). This is in accordance with many studies form developing and developed countries indicating that G. intestinalis is the most common protozoan parasite-causing diarrhea (Thompson et al. 2000; de Wit MAS et al. 2001; Stark et al. 2009; Foronda et al. 2008; El Naggar et al. 2006; Sabry et al. 2009). Comparing microscopy with real-time PCR revealed that microscopy resulted in many false-negative and positive results, giving lower sensitivity, specificity as well as positive and negative predictive values. The lower sensitivity and specificity of microscopy compared to conventional and real-time PCR has already being reported by a number of other studies (Ghosh et al. 2000; Verweij et al. 2003; ten Hove et al. 2007; Calderaro et al. 2010; Stark et al. 2011).

Microscopy showed lower sensitivity of 33.3 % when compared with real-time PCR for detection of Cryptosporidium spp. The parasite was detected in 1 % of the diarrheal patients by microscopy and in 3 % by real-time PCR. Several studies have previously demonstrated that PCR has superior sensitivity for detection of Cryptosporidium spp. when compared to conventional staining and microscopy (Morgan et al. 1998; Kaushik et al. 2008; Stark et al. 2011).

While E. histolytica/dispar was detected in 10.8 % by microscopy. Less than one fifth of them (2 %) were found true positive for E. dispar by real-time PCR. E. histolytica DNA was not detected in any of the diarrheal patients. As E. dispar is nonpathogenic, the presence of this DNA in the stool of diarrheal cases is indicative for an association rather than being the etiologic agent. Comparison between microscopy and real-time PCR for E. histolytica/dispar indicates that microscopy exhibited many false positive results giving a low positive predicative value of 18.6 %. This may be due to misdiagnosis of other Entamoeba species such as Entamoeba coli, Entamoeba hartmanni, or the morphologically identical Entamoeba moshkovskii. (Verweij et al. 2003; González-Ruiz et al. 1994; Tanyuksel and Petri 2003; Hamzah et al. 2010). These findings confirm the limitation of microscopy for the differentiation of the various Entamoeba spp. and calling into question prevalence rates of E. histolytica previously recorded on the basis of microscopy.

In conclusion, the present study has shown that the implementation of multiplex real-time PCR for the simultaneous detection target DNA in a closed-tube system would be beneficial for the rapid and accurate diagnosis of common diarrhea-causing protozoa. For developing countries, although the reagents costs are still high compared to microscopy, multiplex real-time PCR not only simplifies the detection of several enteric pathogens, but also reduces the cost of unnecessary treatment following misdiagnosis. Pooling of samples to a reference laboratory would reduce the running cost of the test.