Introduction

Amoebiasis is still a big challenge to public health in many regions, especially in the ‘bottom billion’ countries where poverty and low income is prevalent, and complex challenges are hindering their economic development. Areas with high rates of amoebic infection include parts of India, Bangladesh, tropical African countries, Brazil and Mexico, China, and South-east Asia [1, 2]. It is estimated to affect 50 million people worldwide and causes up to 100,000 deaths annually [1, 3]. Approximately 90% of infected individuals are asymptomatic carriers; the other 10% show clinical symptoms such as colitis, dysentery and extraintestinal amoebiasis [3]. The most common clinical manifestation of extraintestinal infection is amoebic liver abscess (ALA) and a delay in diagnosis and treatment may cause fatality [4]. Despite the prevalence of amoebiasis, there is still no vaccine to prevent this disease [5]. Human infection is usually found in areas with poor sanitary conditions, inadequate water treatment and low socio-economic status. The only reservoir is human, and infection occurs via food, water or hands contaminated with cyst-containing fecal material. Human to human transmission has been reported through oral-genital and oral-anal contact, especially among homosexuals [6] and those with poor personal hygiene [7].

Diagnosis of intestinal amoebiasis relies on clinical symptoms and laboratory test results. Continuous improvement of health programmes, as well as monitoring and mapping the prevalence of amoebiasis is needed and this requires good diagnostic tools. This review describes the laboratory diagnosis of amoebiasis. Other than the conventional methods, a substantial amount of work has been carried out to develop new and improved serological and molecular diagnostic tests for both clinical and research purposes.

Laboratory diagnosis of amoebiasis

Laboratory diagnostic methods for amoebiasis are based on parasitological, immunological and molecular techniques. The microscopic observation of the parasite in stool, body fluid or tissue sample is considered as the ‘gold standard’ in diagnosis. Patients in endemic areas with clinical signs and symptoms that include gastrointestinal discomfort and watery or bloody diarrhoea should be suspected of intestinal amoebiasis. The laboratory diagnosis of intestinal disease can be made by microscopy, culture, isoenzyme analysis, antigen detection test, molecular-based test and point-of-care (POC) test.

The laboratory diagnosis of extraintestinal amoebiasis is different from intestinal amoebiasis in two ways. First, most patients with extraintestinal amoebiasis, especially ALA, do not have concurrent amoebic colitis. Thus, analysis of stool sample is generally not performed for suspected ALA cases, unless intestinal symptoms are present as well. Second, majority of patients with intestinal amoebiasis have been exposed to Entamoeba histolytica, and developed IgG antibodies to this parasite which may persist for some time. Thus, definitive diagnosis using the available IgG antibody detection assays is a challenge because of the difficulty in differentiating past and current infections [8, 9].

Intestinal amoebiasis

Microscopic examination

The visual demonstration of E. histolytica cysts and/or trophozoites in stool or colonic mucosa of patients can be performed by microscopic examination. This technique is still frequently practiced in many parasitology diagnostic laboratories, particularly in developing countries [3]. The direct examination of a saline wet mount of fresh sample (with or without iodine as temporary stain) under a microscope is not a sensitive method [10]. To identify motile trophozoites (which may contain red blood cells), the stool samples need to be examined within 1 h of collection. Therefore, if the examination cannot be performed immediately, the stool sample should be preserved in polyvinyl alcohol (PVA), Schaudinn’s fixative or sodium acetate-acetic acid-formalin (SAF) [11]. The possibility of observing trophozoites is higher in loose stools, which contain mucous, pus and trace amounts of occult blood, whereas cysts can be observed in both formed and loose stools [12]. Permanent stain of the stool smear should be examined to enable the morphology, size and number of the nuclei to be clearly observed. Stains such as methylene blue, Giemsa, Wright’s and iodine-trichrome can be used for the staining, however for routine use, the modified iron haematoxylin and Wheatley’s trichrome stains are recommended [3]. Even though microscopic examination allows visualization of the parasite and hence provides a definitive diagnosis, it has several limitations. Before the morphological similar non-pathogenic strain E. dispar was discovered, misdiagnosis and over-treatment were common. The morphologies of E. histolytica, E. dispar and E. moshkovskii under the microscope are indistinguishable, although the presence of ingested red blood cells most likely indicates infection with E. histolytica. Moreover, although these three species can be differentiated morphologically from the other common amoebas (E. coli, E. hartmanni, E. polecki and I. butschlii), it is still a challenge for an inexperienced technician. Thus, the diagnostic sensitivity and specificity of microscopic examination to detect E. histolytica in stool is considered low [13,14,15,16,17].

Biochemical methods: culture and isoenzyme analysis

Previously, stool culture followed by isoenzyme analysis was commonly used as a gold standard method to differentiate between E. histolytica and E. dispar [18]. Other than faecal sample, rectal biopsy or liver abscess aspirate can also be used to culture E. histolytica. Pus aspirate from liver of an ALA patient is normally sterile thus it is necessary to add a bacterium or a trypanosomatid before introducing the amoebae into a xenic culture [18,19,20].

From the cultured amoeba, isoenzyme analysis is performed using zymodeme enzymes as markers [21]. A zymodeme is a cluster of amoeba strains that has the same electrophoretic pattern and mobility for a few enzymes. Examples of the enzymes are hexokinase, decarboxylating malate dehydrogenase, glucose phosphate isomerase and phosphoglucomutase isoenzyme [22]. There are 24 different established zymodemes in which 21 are from human isolates (nine E. histolytica and 12 E. dispar) and another three zymodemes from experimentally cultured amoeba strains. Since E. histolytica and E. dispar have genetically different hexokinase enzymes, it is reliable in discriminating between the two species. Three zymodeme bands for E. histolytica (II, XIV, and XIX) as compared to only one band for E. dispar (I) can be used to differentiate the two Entamoeba species [23].

However, isoenzyme analysis requires the use of cultured amoeba trophozoites which is tedious and time consuming [12, 24,25,26]. Four to 10 days are needed to grow the trophozoites to a significant amount prior to performing starch-gel electrophoresis, and the culture may not be always successful [16]. In reference laboratories, the success rate of establishing E. histolytica culture was reported to be between 50 and 70% [18]. The isoenzyme analysis of E. histolytica culture from clinical samples often gives false-negative result. There were also many samples that were positive by microscopy but were culture-negative [27]. In addition, a major problem that may arise during E. histolytica culture is the overgrowth of bacteria, other protozoan or fungi [18]. Therefore, due to its low sensitivity, culture in combination with isoenzyme analysis, is not routinely used in diagnosis [28]. This technique is more suitable for research rather than as a primary diagnostic tool. Molecular diagnosis has now replaced isoenzyme analysis as the preferred method to identify Entamoeba species.

Antigen detection ELISA

The disadvantages of the traditional parasitological techniques have led to the current use of coproantigen ELISAs for laboratory diagnosis of intestinal amoebiasis. ELISAs are useful for clinical and epidemiological studies, especially where molecular assays are not practical or available [25, 26]. The immunoassay is relatively simple and rapid, and can be performed in most laboratories. TechLab E. histolytica II ELISA (TechLab, Blacksburg, VA, USA) and Entamoeba CELISA PATH kit (Cellabs, Brookvale, Australia) use monoclonal antibodies against E. histolytica Gal/GalNAc lectin. Other commercial ELISA kits include Optimum S kit (Merlin Diagnostika, Bernheim-Hersel, Germany) and ProSpecT ELISA (Remel Inc., Lenexa, KS, USA). The former detects serine-rich antigen of E. histolytica while the latter detects a specific antigen (EHSA) from E. histolytica/E. dispar [15]. Comparison of diagnostic sensitivity and specificity of these kits obtained from different studies are presented in Table 1. The most commonly used antigen detection test is the E. histolytica TechLab kit. It is the first generation kit in ELISA format produced in 1993 to specifically detect E. histolytica Gal/GalNAc lectin in stool samples [26, 27]. This lectin protein is highly immunogenic and conserved, and can be used to specifically detect E. histolytica due to the antigenic differences between the lectins of E. histolytica and E. dispar. According to Haque et al. [27], this test showed an excellent correlation with nested PCR when tested with stool samples from people with diarrhoea. Moreover, this test was reported to have higher sensitivity (80 to 94%) and specificity (94 to 100%), as compared to both microscopy and culture [13, 25]. However, in a study by Gonin and Trudel [55], E. histolytica TechLab kit was found to show reduced diagnostic sensitivity and specificity compared to microscopy and PCR in discriminating between E. histolytica and E. dispar.

Table 1 Commercially available antigen and antibody detection assays, and rapid diagnostic tests for intestinal and extraintestinal amoebiasis
Full size table

Due to some limitations observed in the first generation TechLab ELISA kit, a second version of the kit called TechLab E. histolytica II was produced. In a study performed in Bangladesh, it was reported to display higher sensitivity (86% to 95%) and specificity (93% to 100%) when compared to microscopy and culture [13, 28]. It also demonstrated good levels of sensitivity (71 to 79%) and specificity (96 to 100%) when compared to real-time PCR for the diagnosis of E. histolytica [30, 31, 55]. In addition, Haque et al. [56] also used this kit on serum and liver abscess samples of patients. The result showed that, prior to treatment with metronidazole, 96% (22/23) and 100% (3/3) of the ALA patients can be detected to have the lectin antigen in their serum and liver abscess samples, respectively. However, after several days of the treatment, the diagnostic sensitivity decreased to 33% (32/98) and 41% (11/27) for serum and liver abscess, respectively. This is likely due to the reduction of lectin in the samples post-therapy. In contrast, another study which was performed in the village of Borbòn, Ecuador, reported that TechLab II E. histolytica showed low diagnostic sensitivity (14.3%) for detection of E. histolytica antigen when compared to culture and zymodeme analysis [29]. Visser et al. [31] also found that the kit lacked sensitivity for a reliable diagnosis of E. histolytica infection among carriers of the parasite in a non-endemic area. In other studies conducted in North India and Baghdad, Iraq, the sensitivity and specificity of this test was reported to be 20–60% and 86.7–93.4%, respectively, when compared with microscopy-positive E. histolytica, E. dispar, or E. moshkovskii [32, 33].

ProSpecT ELISA is a microplate immunoassay assay that detects both E. histolytica and E. dispar antigens. In comparison with conventional microscopy, the sensitivity and specificity of this test was reported to be 78% and 99%, respectively [35]. In another study by Gatti et al. [29], this assay was found to be 54.5% sensitive and 94% specific for detection of E. histolytica/E. dispar as compared to culture and zymodeme analysis. In Australia, Stark and colleagues [57] evaluated the use of the CELISA PATH and TechLab E. histolytica II kits to detect E. histolytica using PCR as the reference standard. Both kits use monoclonal antibody against the Gal/GalNAc lectin of E. histolytica. The former showed 28% sensitivity and 100% specificity, while the latter showed very low sensitivity and specificity. The TechLab E. histolytica II kit required 10,000 trophozoites/well for positive result, hence less sensitive as compared to Entamoeba CELISA PATH kit, which required approximately 1 000 trophozoites/well. It was suggested that the different amounts of antibody used to coat the wells of the plates might contribute to the differences in performance between the two ELISAs [58]. A recent study on 288 stool samples of children in a community village in Budhni, Peshawar, using Entamoeba CELISA PATH showed diagnostic sensitivity and specificity of 27% and 98.4% respectively, compared to microscopy-positive Entamoeba species [34]. Meanwhile, Pillai et al. [36] evaluated the usefulness of Optimum S kit to detect E. histolytica in 72 stool samples positive for E. histolytica/E. dispar complex; it showed only 4.2% sensitivity as compared to the combined results of two other coproantigens ELISAs.

Besides the commercially available kits, several laboratory based-assays have been developed using monoclonal and polyclonal antibodies against various E. histolytica antigens such as lipophosphoglycan, lectin-rich surface antigen and pyruvate phosphate dikinase [PPDK] [54, 59, 60]. In addition, a 170-kDa amoebic adherence lectin was reportedly detected in saliva of amoebiasis patients [61].

Molecular diagnosis

Molecular diagnostic tests have gained primacy for detection of various infectious diseases, including amoebiasis [62]. There are different variants of DNA amplification techniques for differentiation and detection of the Entamoeba species in stools, tissues and liver lesion aspirates. They include conventional PCR, nested PCR, real-time PCR, multiplex PCR and loop-mediated isothermal amplification assay (LAMP) [63,64,65,66,67]. Table 2 shows the various assay types and parameters used in molecular diagnosis of amoebiasis.

Table 2 Assay types and parameters used in molecular diagnosis of intestinal and extraintestinal amoebiasis
Full size table

Conventional PCR

To date, many genes are targeted for recognition and discrimination of the three Entamoeba species, i.e. small subunit rRNA, gene encoding a 30-kDa protein, DNA highly repetitive sequences, haemolysin gene (HLY6), cysteine proteinase, serine-rich E. histolytica (SREHP) gene, actin gene and tandem repeats in extrachromosomal circular DNA [68, 69, 81, 91, 92]. The HLY6 gene has been used to develop a PCR assay for the detection of E. histolytica DNA in stool samples, and showed 100% diagnostic sensitivity and specificity [69]. However, PCR assay targeting small subunit rRNA is most commonly used due to its presence in multiple copies of the extrachromosomal plasmids [93]. In addition, due to a high genetic variation between 18S rRNA gene of E. histolytica and E. dispar, it can be used to differentiate between these two species [92, 94].

A PCR targeting the small subunit rRNA gene has been developed to detect Entamoeba species in stool samples [70]. Seven out of 27 microscopy-positive stool samples were successfully identified by PCR in a Thai population. In Indonesia, the same gene target was used in a multiplex PCR performed on 30 samples of diarrheic stools. It showed 12 positive results, from which seven were positive for E. histolytica, two for E. moshkovskii, and three showed mixed infection of E. histolytica and E. moshkovskii, and no positive result for E. dispar [71]. A molecular epidemiology study among North East Indian population showed that the overall prevalence with any of the three morphologically indistinguishable Entamoeba species was 23.2% (95% CI = 20.9%, 25.6%). Of these, 13.7% (173/1260; 95% CI = 11.9, 15.7) and 11.8% (149/1260; 95% CI = 10.2, 13.8) of the subjects were infected with E. histolytica and E. dispar, respectively. The former group was PCR-positive either singly for E. histolytica or in combination with other intestinal protozoan parasites [95].

Nested PCR

Two nested PCR assays on DNA of stool samples targeting a 16S-like rRNA were reported by the International Centre for Diarrheal Diseases and Research, Dhaka, Bangladesh [72] and Jawaharlal Institute of Postgraduate Medical Education & Research, Puducherry, India [73]. They gave 100% diagnostic specificity and were able to distinguish infections caused by E. histolytica, E. dispar and E. moshkovskii. In a prevalence study in Malaysia, the nested PCR successfully detected 75% E. histolytica, followed by E. dispar (30.8%) and E. moshkovskii (5.8%) [74]. Another study on aborigines in Malaysia showed that 52 (80%, n = 65) microscopically positive samples were successfully amplified by the nested PCR. Of these, 65.4% were found to be E. histolytica and 13.5% were E. dispar and none was positive for E. moshkovskii. The nested PCR was also 100% specific as no amplification of other genomic DNA was observed [75]. In Mexico, a nested PCR was designed to detect and differentiate E. histolytica from E. dispar using a fragment of the adh112 gene. Of 62 samples tested, 16.1% were positive for E. histolytica while none was positive for E. dispar [76]. Nevertheless, the identification of the Entamoeba species from stool specimens by nested PCR for individual species is a tedious process.

Nested multiplex PCR

In a trial to increase the sensitivity of PCR method, a nested multiplex PCR was developed for simultaneous detection of Entamoeba species using DNA extracted from stool specimens. Khairnar and Parija [96] found that the assay showed 94% sensitivity and 100% specificity. Fallah et al. [77] reported that the nested multiplex PCR was useful for the specific detection of pathogenic and non-pathogenic Entamoeba species in stool samples. Thirty-one (4.28%) out of 724 stool samples were positive for E. histolytica/E. dispar, with 54.8% samples positive for E. dispar and 8 (25.8%) samples positive for E. histolytica. In Egypt, a nested multiplex PCR detected samples positive of E. histolytica and E. dispar with a sensitivity of 96.8%. Of these samples, 17 (32.7%) were positive for E. histolytica, 12 (23.1%) for E. dispar, and 3 (5.7%) for both species. However, the nested multiplex PCR detected E. dispar in one of the negative control samples, thus giving a diagnostic specificity of 95%. In a study by ElBakri et al. [97], the nested PCR was used for simultaneous detection of E. histolytica, E. dispar and E. moshkovskii from 120 faecal samples collected from Sharjah Emirate, UAE. The result showed that 10% (12/120) samples were mono infected with E. histolytica; 2.5% (3/120) with E. dispar; and 2.5% (3/120) with E. moshkovskii. Furthermore, mixed infections by both E. histolytica and E. dispar were observed in 3.3% (4/120) samples; and E. dispar and E. moshkovskii in 0.8% (1/120) samples. Meanwhile, in Nigeria, the PCR results showed that out of 46 microscopy-positive samples, 16 (34.8%) successfully generated species-specific amplicons of Entamoeba species DNA. Infection with E. dispar (68.8%; 11/46) was the most common, followed by E. histolytica (37.5%; 6/46) and E. moshkovskii (18.8%; 3/46) [78]. A first molecular epidemiology study of E. histolytica, E. dispar and E. moshkovskii infection in Yemen using multiplex nested PCR revealed that 20.2%, 15.7% and 18.2% of 605 samples were positive for E. histolytica, E. dispar and E. moshkovskii, respectively [79]. These result suggested that nested multiplex PCR offers high diagnostic sensitivity and specificity, and allows simultaneous detection and differentiation between E. histolytica and E. dispar in microscopy-positive stool samples [70, 96].

Real-time PCR

Conventional and nested PCR are time consuming especially when processing many samples, costly, results are non-quantitative, and may produce false positive results due to carry-over contamination [3, 15]. Thus, real-time PCR assay has gained a lot of attention for laboratory diagnosis of infection since it can enhance diagnostic sensitivity, eliminate post-PCR manipulation and minimize contamination [3]. It also allows quantification of the relative number of parasites present in various forms of clinical samples including stool, liver abscess, aspirate and urine [96]. Many studies have been published on the use of real-time PCR assays for specific detection of E. histolytica and E. dispar using DNA from stool samples [30, 66, 80, 81, 98]. Most of them targeted either 18S rRNA gene or species-specific episomal DNA repeat genes. For example, Qvarnstrom et al. [80] compared real-time PCR using different probes, and the results suggested that the TaqMan method, which targets the 18S rRNA gene, was more specific than the SYBR Green approach for diagnosis of amoebiasis. They also reported that the probe-based real-time PCR assays can be used to detect E. histolytica in clinical samples with very low number of parasites which are not detectable by conventional PCR. In Mexico, analysis of stool samples from 273 children using Faust stool concentration technique showed that 25 (9.2%) were positive for E. histolytica/E. dispar/E. moshkovskii. Of these, 3 were positive for E. histolytica by SYBR Green real-time PCR; and 2 samples that were negative for E. histolytica/E. dispar/E. moshkovskii by the Faust technique were positive by the real-time PCR [85]. Meanwhile, in another study, 672 stool samples from endemic areas in Vietnam and South Africa were used to evaluate a real-time PCR targeting 310 bp fragment of rDNA-containing amoeba episome. The results were compared with those from amoeba culture and microscopy [66]. They reported that all samples positive by microscopy and 88% samples positive by culture were also positive by real-time PCR. When compared to culture and isoenzyme analysis, the real-time PCR was 100% specific for detection of both Entamoeba species. Roy et al. [30] showed that real-time PCR was 99% sensitive and specific for detection of E. histolytica in stool as compared to conventional PCR. Another recent study from Egypt used nested multiplex PCR and TaqMan real-time PCR to determine the prevalence of E. histolytica. Among 194 microscopy-positive Entamoeba samples, the nested PCR identified 8.7% (n = 17) as E. dispar and 10.3% (20/194) as E. histolytica. With the real-time PCR, 5.7% (11/194) and 9.8% (19/194) samples were found to be positive for E. dispar and E. histolytica respectively [82].

Multiplex real-time PCR

Recent developments in multiplex real time PCR make it possible to rapidly identify, genotype and quantify multiple DNA targets simultaneously in a single reaction. In Thailand, a multiplex real-time PCR was established to differentially detect E. histolytica, E. dispar and E. moshkovskii [86]. The assay detected E. histolytica in four of 32 microscopy-positive stool samples. Most of the stool samples were positive for E. dispar, and one sample had mixed infection with E. moshkovskii. Meanwhile, Liang et al. [58] reported the use of a single-tube multiprobe real-time PCR (EntaTaq) assay for simultaneous detection of E. histolytica and E. dispar. The assay identified 23.5% (12/51) E. histolytica and 41% (16/39) E. dispar in samples which were negative by nested PCR, without cross reactivity with other commensal protozoa. Based on the data of previous studies, the detection limit of the EntaTaq was 10 times greater than nested PCR (10–100 trophozoites/ml) [17]. In another study, stool samples were analyzed from 396 Egyptian patients with diarrhoea, along with 202 samples from healthy controls. A total of 43 patient samples were microscopy-positive for E. histolytica/dispar; however, a real-time PCR assay only detected eight samples with E. dispar while E. histolytica was not detected at all [83]. Thus, the use of the real-time PCR for simultaneous detection of multiple DNA targets would be beneficial for the accurate and rapid diagnosis of amoebiasis.

Currently, several multiplex PCR panels have been certified as in vitro diagnostic tests (IVD), allowing the simultaneous detection of multiple pathogens in stool samples [99, 100]. Table 3 illustrates the performances of commercially available rapid PCR-based tests for the detection of E. histolytica in stool samples. In one study using real-time multiplex PCR, the RIDA GENE Parasite stool panel II (R-Biopharm, Darmstadt, Germany) was evaluated for the specific detection of E. histolytica, Giardia lamblia and Cryptosporidium parvum in 180 patients suffering from diarrhoea who attended the outpatient clinic of a central teaching hospital for children in Baghdad city, Iraq. These results showed 100% sensitivity and specificity when compared to microscopy for the detection of E. histolytica [101]. A study conducted by Laude et al. [105] using 185 stool samples collected from 12 parasitology laboratories in France reported that the G-DiaParaTrio® (Diagenode Diagnostics, Belgium) multiplex PCR assay identified 38 samples with G. intestinalis, 25 with C. parvum/C. hominis and 5 with E. histolytic, leading to sensitivity/specificity of 92%/100%, 96%/100% and 100%/100% for G. intestinalis, C. parvum/C. hominis and E. histolytica, respectively. A recent study conducted on 172 clinical stool samples using multiplex PCR ParaGENIE G-Amoeba (Ademtech, France) reported a high sensitivity and specificity for the detection of G. intestinalis and E. dispar/E. moshkovskii from stools (89.7%/96.9% and 95%/100%, respectively). Although two false-negative samples were observed which were confirmed to be positive for E. histolytica by an external PCR assay, the multiplex PCR showed better specificity compared to an antigen detection test (ELISA) [104]. The Luminex xTAG gastrointestinal pathogen panel assay (GPP) (Luminex Molecular Diagnostics, Austin, TX, USA) is another commercial qualitative multiplex test, it able to identify 19 enteric pathogens in one reaction in 6 h. The test sensitivity and specificity for detection of E. histolytica from clinical specimens was reported to be 100% and 89%, respectively as compared to real time PCR [103]. Although real-time PCR assay is highly sensitive and specific, the major disadvantage of this method is its relative high cost, thus may not be utilized by most laboratories in developing countries [106].

Table 3 Performances of commercially available rapid PCR-based tests for detection of Entamoeba histolytica in stool specimens
Full size table

Loop mediated isothermal amplification assay (LAMP)

The use of isothermal amplification assay is a good choice for molecular diagnosis especially for low resource areas due to its high sensitivity, specificity, rapidity and simplicity [107, 108]. Liang et al. [17] have developed a loop-mediated isothermal amplification assay (LAMP) assay for intestinal amoebiasis and compared the results with a nested PCR. E. histolytica DNA was detected in 60% samples (18/30) using LAMP and 33% samples (10/33) using nested PCR, thus showing the superior sensitivity of the former. In a study conducted by Singh et al. [87], the LAMP assay was found to be more sensitive than conventional PCR assay for diagnosis of ALA. Out of 50 pus samples, 36 (72%) were positive for E. histolytica by conventional PCR assay and 41 (82%) by LAMP assay. However, a study by Rivera and Ong, [88] in the Philippines found that the detection limit of the LAMP assay was five parasites per reaction, which corresponded to approximately 15.8 ng/μl DNA, while the detection limit for conventional nested PCR was 2 ng/μl DNA samples. A modification of the isothermal assay based on 18S small subunit ribosomal RNA gene and designed with extra reaction accelerating primers is called stem LAMP. Mwendwa et al. [89] used the stem LAMP and detected 36 of 126 (28.6%) DNA samples as E. histolytica, and this was better when compared to the standard LAMP test (15.9%, 20/126) or conventional PCR (13.5%, 17/126). In another study, a four target nitrocellulose-based nucleic acid test using lateral flow immunoassay biosensor that detected amplicons from a thermostabilized triplex LAMP assay was developed. The biosensor detection limit was 10 E. histolytica trophozoites and showed 100% specificity when evaluated against three medically important Entamoeba species and 75 other pathogenic microorganisms [90].

Rapid diagnostic test

In recent years, there is emergence of rapid diagnostic tests as point-of-care (POC) tests for amoebiasis [109]. Such a test represents a bridge that can connect the laboratory and the field. It allows for the mass screening of endemic populations and for the monitoring of control programmes, thereby providing quick diagnosis and reducing disease transmission. It also provides rapid result to a physician, thus shortening the duration gap between disease diagnosis and treatment [110]. As compared to the other laboratory tests such as ELISA and PCR, rapid test avoids the need of expensive equipment. Thus, it is a preferred diagnostic tool in developing countries with low resources [111].

There are a number of rapid diagnostic tests available in the market for intestinal amoebiasis (Table 1). The Triage Parasite Panel (TPP) (Biosite Diagnostic Inc., San Diego, CA, USA) is the first lateral flow test to simultaneously detect antigens specific for G.lamblia, E. histolytica/E. dispar and C. parvum. The test uses monoclonal antibodies specific to a 29-kDa surface antigen of E. histolytica/E. dispar, G. lamblia alpha-1-giardin, and C. parvum protein disulfide isomerase. Two studies showed that the TPP kit had high diagnostic sensitivity (96% to 100%) and specificity (99.1% to 100%) for detection of E. histolytica/E. dispar, when compared to microscopy [45,46,47]. In contrast, Garcia et al. [45] and Pillai and Kain [36] found that the sensitivity of the kit was low (68.3%), albeit with high specificity (100%) when compared with ProSpecT test. This is corroborated by the work by Leiva et al. [112], who found that the sensitivity of TPP kit was low when compared to PCR assay. In addition the kit is unable to differentiate among E. histolytica, E. dispar and E. moshkovskii. Furthermore, either fresh or fresh-frozen non-preserved stools should be used with the TPP kit which may be impractical in some situations [3].

A retrospective study was performed by Van den Bossche and his colleagues [48] to evaluate the lateral flow RIDA®QUICK Cryptosporidium/Giardia/Entamoeba Combi (R-BioPharm, Darmstadt, Germany), using stool samples collected from patients at an outpatient clinic of the Institute of Tropical Medicine (ITM), Antwerp, Belgium or the Central Laboratory of Clinical Biology (CLKB). The kit demonstrated 100% sensitivity, while the specificity ranged from 80% to 88% for detection of E. histolytica. This result differed from Goñi et al. [49], which showed that the kit exhibited 62% sensitivity and 96% specificity for E. histolytica. The lower specificity in the study by Van den Bossche et al. [48] can be explained by the high number of E. dispar samples, which substantially influences specificity since this kit is unable to differentiate between E. histolytica and E. dispar [15]. Another version of the one-step immunochromatographic test for the qualitative detection of C. parvum, G. lamblia and E. histolytica antigens in human stool samples has been introduced to the market by Meridian Bioscience Inc. (Luckenwalde, Germany) and is known as ImmunoCard STAT!® CGE. This new ImmunoCard rapid antigen detection test exhibited 88% sensitivity and 92% specificity as compared to real-time PCR in detection of E. histolytica, but it showed cross-reactivity with E. dispar [50]. A study was conducted on diarrheic/dysenteric stool samples from clinically suspected individuals from Beni-Suef, Egypt, using RIDA®QUICK Entamoeba Test (R-Biopharm AG, Darmstadt, Germany), an immunochromatographic (ICT) rapid assay for the qualitative determination of E. histolytica / dispar, and Techlab E. histolytica II ELISA test was used as the reference. Of 7 specimens that were positive by the ELISA, only 2 specimens were positive by the ICT, thus a sensitivity of 28.6%, and the specificity was reported as 86.1% [51].

The third generation of a rapid test known as E. histolytica Quik Chek (TechLab, Blacksburg, VA, USA) was recently introduced to the market. The antibody used in this kit is specific against the E. histolytica adherence lectin [113]. The test is a modified version of the TechLab E. histolytica II ELISA, but uses a flow-through format. An evaluation of this point-of-care rapid test in a cohort of children in Bangladesh showed 100% sensitivity and specificity when compared to an ELISA antigen detection assay [52]. When compared to ProSpecT microplate assay, this kit exhibited 97% sensitivity and 100% specificity [53]. These findings indicated that the new Quik Chek assay was robust and can specifically detect E. histolytica trophozoites in unfixed, frozen clinical stool samples. However, it requires an additional incubation of both conjugate and chromogen which increases the processing time. Furthermore, it requires cold-chain transportation.

Our group previously reported the development of a lateral flow dipstick test which detected E. histolytica PPDK in stool samples [54]. When compared to real-time PCR, the diagnostic sensitivity of the dipstick was 65.4% (n = 17/26), while the specificity when tested with stool samples containing other intestinal pathogens was 92% (23/25). Although not highly sensitive, it was superior to the performance of Techlab E. histolytica II ELISA which detected only 19.2% (5/26) of the same set of PCR-positive samples. Thus, the lateral flow dipstick test showed good potential to be further developed into a stool rapid test for intestinal amoebiasis.

Extraintestinal amoebiasis

Clinical manifestations of ALA are highly variable, thus making the diagnosis difficult [56, 114]. Ultrasound, computed tomography and magnetic resonance are very useful techniques and have excellent sensitivity for detection of liver abscess arising from any cause [115]. However, these detection methods cannot distinguish amoebic abscesses from pyogenic abscesses or necrotic tumours. Most patients with ALA do not have co-existing amoebic colitis. Only less than 10% of amoeba have been identified in stool samples from ALA patients using microscopy and antigen detection tests [116]. Consequently, stool microscopy or antigen detection tests are not helpful for diagnosis of ALA [27, 56, 61].

Microscopic examination

ALA can be confirmed by microscopic examination of liver pus to look for amoebic trophozoites [117]. However, aspiration of liver pus in an ALA patient is not necessary for establishing the diagnosis. When aspirated, they contain acellular debris that forms a brown, thick fluid (‘anchovy paste’). Although the presence of amoebic trophozoites in aspirated pus is confirmatory for ALA, it is not a sensitive diagnostic method. Trophozoites are seen in a minority of aspirates (< 20%) and typically only seen when the wall of the cyst is sampled [118, 119]. Hence, it is not surprising that many authors reported their total absence or very low incidence on examination of liver pus. In general, trophozoites have been seen in aspirated pus in 11–25% cases [120]. Parija and Khairnar [121] found trophozoites of E. histolytica in pus in 7.2% cases, while Haque et al. [56] showed that 11% of aspirated liver pus samples were positive for the organism.

Microscopic examination of the pus may show dead and deformed hepatocytes, red blood cells and some polymorphs. A good staining method is necessary to visualize the morphological changes in the liver tissue and also differentiates the amoebas against the surrounding cells. The common staining techniques are haematoxylin and eosin (H&E), periodic-acid Schiff (PAS) and immunostaining. With H&E and PAS stains, amoebic trophozoites are difficult to differentiate from macrophages because of similarities in size and morphology [122]. Examination of fixed and stained biopsy samples using H&E stain is challenging due to lack of differentiation between the stained trophozoites and surrounding tissues [123]. With PAS stain, the amoebae are intensely PAS positive due to the presence of glycogen in cytoplasm; however, liver cells also contain glycogen [124]. Moreover, these two staining methods require high technical expertise to identify and interpret the results. In comparison, immunohistochemistry (IHC) is presumed to be more specific as it is the consequence of specific reactions between amoebic antigens and antibodies against them. Although IHC is still rarely reported for use in investigation of invasive amoebiasis [122], it is potentially very useful as a confirmatory test for ALA when liver pus aspirate biopsy or autopsy is available.

Antibody detection test

Petri and Singh [60] reported that anti-amoebic antibodies were detectable in more than 90% of patients with ALA within 7 to 10 days after symptoms began [117]. Therefore, serology is used for diagnosis of ALA, in combination with observations from clinical manifestations and result of radiological imaging. Several assays are commonly used to detect the anti-amoebic antibodies, such as ELISA, indirect haemagglutination (IHA), indirect immunofluorescence assay (IFA), latex agglutination, immunoelectrophoresis, counterimmunoelectrophoresis (CIE), amoebic gel diffusion test, immunodiffusion and complement fixation test [125]. Of these, ELISA is the most popular technique and has been used to investigate the epidemiology of symptomatic amoebiasis due to its reliability and ease of performance [113, 126].

A study was performed to compare the performance of several antibody detection assays on patients in Kuwait suspected of ALA using IHA (Collegnost Amoebiasis, Germany), Amoebiasis Serology Microwell II EIA (LMD LAboratories, CA) and ImmunoTab test (Institute Virion, Germany). The true positive cases were defined as patients positive by CT scan or ultrasonography with an IHA titer ≥ 1:256. The sensitivity and specificity of the IHA, ImmunoTab and Microwell II EIA was found to be 99% and 99.8%, 98% and 95.5% and 97.9% and 94.8%, respectively [37]. In other studies, the sensitivity of the Microwell II EIA for the detection of specific E. histolytica antibodies in ALA patients was reported to be 97.6–99% [125, 127].

Hung et al. [128] reported that IHA showed high specificity (99.1%) when tested with samples from patients with gastrointestinal symptoms, including HIV-infected patients. However, it showed lower sensitivity when compared to ELISA due to false-negative results [129]. In a village in West Kalimantan, Borneo, the seroprevalence of anti-amoebic antibodies among the population with IHA titers equal or greater than 1:128 was found to be 7% [130]. Meanwhile, the seroprevalence among blood donors from Kelantan, Malaysia, was reported to be 16% [131]. A study conducted by Dhanalakshmi et al. [43] using IHA, based on a recombinant calcium binding domain-containing protein, on serum of suspected amoebiasis patients showed 62% sensitivity and 96% specificity. In comparison, an antigen detection ELISA on the same set of samples showed 69% sensitivity and 90% specificity.

Garcia et al. [132] reported that the IFA was a reliable, reproducible and a rapid tool to differentiate ALA from other non-amoebic etiologies. In addition, it helps to differentiate current and past (treated) infected patients. The sensitivity of IFA was reported to be 93.6% (higher than ELISA) and the specificity was 96.7% [133]. However, this technique is difficult to perform routinely since it requires skill to culture the parasite and prepare the antigen [134].

The common commercially available antibody assays for detection of E. histolytica antibodies in human serum are included in Table 1. Most of these commercial tests are coated with crude soluble trophozoite antigen for the detection of anti-amoebic antibodies. Several studies have demonstrated high sensitivity of crude soluble and excretory-secretory antigens in detecting amoebic antibodies in ALA patients [38,39,40, 42, 135, 136, 137]. However, the drawbacks using crude antigens are the high cost and tediousness of maintaining E. histolytica culture, and other problems associated with the mass production of the antigen. In addition, the crude amoebic antigen gave false positive results, thus decreasing the test specificity [138]. In endemic settings, it tends to produce high background readings; thus, the antibody cut-off titre may be needed to be adjusted from that recommended by the manufacturer. In general, the use of the current commercial antibody detection tests is only reliable in developed countries where amoebiasis is not endemic. However, if the vast majority of patients in these countries are immigrants from developing countries, the usefulness of this test in developed countries may also be doubtful [60]. In endemic areas, individuals are constantly exposed to E. histolytica, and therefore the available IgG antibody detection tests are unable to definitively distinguish between past and current infections [27, 139]. The test may remain positive in patients after a few years of infection, and thus is a major problem that must be addressed. To overcome this limitation, a standardized serological test based on a well-defined antigen is required. A heavy subunit of E. histolytica lectin (152 kDa) and PPDK (110 kDa) for serodiagnosis of ALA have been shown to possess diagnostic sensitivities of above 80% and no cross-reactivity in Western blot analysis [137]. The recombinant form of the PPDK has been produced and Western blot using a panel of serum samples probed with horseradish conjugated anti-human IgG4 showed a high diagnostic sensitivity (93.3%) and specificity (100%), when compared to blots using IgG and IgG1 as secondary antibodies [140].

Other recombinant proteins that has been used as antigens in amoebiasis serology included serine rich protein, 170 kDa subunit galactose-specific adhesin, cysteine proteinase, putative alcohol dehydrogenase and phosphoglucomutase [141,142,143, 144, 145].

Molecular diagnosis

Table 2 includes information on the assays that have been used in detecting E. histolytica DNA in liver abscess/pus samples. Detection of ALA by conventional PCR showed varied diagnostic sensitivity, ranging from 33% to 100% [56, 68, 121, 125, 146]. When PCR using specific primers for E. histolytica 18S rDNA was used to detect the parasite in liver pus samples, only 33% sensitivity was recorded; whereas 100% sensitivity was obtained when a second primer pair specific for a gene encoding a 30-kDa antigen was used [125, 147]. Zaman et al. [68] was successful in obtaining 100% diagnostic sensitivity by using two pairs of published primers namely P1-P2 and P11-P12 which targeted the extrachromosomal circular DNA of E. histolytica and the 30-kDa antigen gene, respectively. In another study, PCR was performed which targeted E. histolytica HLY6 gene using pus samples from ALA patients. The hemo-PCR detected 89% of the samples, as compared to 77% and 28% detection when PCR assays were performed using primers based on the 30-kDa antigen gene and 18S rDNA, respectively [69]. In Canada, a total of 25 ALA samples were assessed by microscopy, serology, and nested PCR. The results showed that microscopy was negative for E. histolytica in all samples, and nested PCR as well as serology were positive in 11 samples [148].

In another study, Parija and Khairnar [121] reported that a nested multiplex PCR detected E. histolytica DNA in 100% (n = 37) liver abscess pus samples from ALA patients collected prior to metronidazole treatment, as compared to only 70.6% (53/75) after therapy. The decline in diagnostic sensitivity can be attributed to the parasite DNA clearance from the abscess subsequent to the parasite destruction post-treatment. However, pus aspiration carries the risk of bacterial infection or spillage of abscess content; thus, it is performed only if there is a danger of rupture of the abscess or if the size of abscess is greater than10 cm in diameter. Abscess should also be drained if it does not respond to medical therapy within 3 to 5 days [149, 150]. In addition, it was reported that the PCR detected E. histolytica DNA in 21 of 53 (39.6%) urine specimens of ALA patients. Among them, four of 23 (17.4%) urine specimens were collected prior to metronidazole treatment, and 17 of 30 (56.7%) were collected after treatment [121]. A study conducted in Malaysia showed that real-time PCR successfully detected E. histolytica DNA in 76.7% of 30 liver abscess samples, whereas IHA detected the presence of anti-E. histolytica antibodies in 46.7%(14/30) of the corresponding serum samples [56]. Meanwhile, a study conducted in Bangladesh by Haque et al. [63], using real-time PCR on 98 ALA patients showed that the assay was able to detect E. histolytica DNA in 49%, 77% and 69% of blood, urine and saliva samples, respectively. Ahmad et al. [84] showed that E. histolytica detection by real-time PCR had higher diagnostic sensitivity than antigen detection method. In a study by Roy et al. [30], a real-time PCR assay on liver pus samples utilizing a molecular-beacon probe was found to be comparable to the TaqMan-based method by Blessmann et al. [66].

Weitzel et al. [102] reported the first use of Rida®Gene Parasitic Stool Panel (R-Biopharm, Darmstadt, Germany) and FilmArray Gastrointestinal Panel (BioFire Diagnostics, Salt Lake City, UT, USA) on samples from aspirated pus of patients with cystic focal liver lesions. Both commercial kits confirmed the diagnosis of ALA within a short time, with FilmArray system giving a faster result than the Rida®Gene test (1 h vs. 3 h) and required less hands-on time (5 min vs. 45 min).

Rapid diagnostic test

To date, there is no commercial rapid diagnostic test for the diagnosis of ALA. However, a proof-of-concept of a lateral flow dipstick test for rapid detection of ALA has been developed. The test is based on detection of anti-PPDK IgG4 antibody in infected patients. Initial evaluation of the rapid test showed 87% diagnostic sensitivity and 100% specificity [140]. Recently, an immunochromatographic test using fluorescent silica nanoparticles coated with C-terminal region of the intermediate subunit of E. histolytica Gal/GalNAc lectin protein has been reported. The kit showed 100% sensitivity and 97.6% specificity when tested with sera from healthy controls and patients with other infectious diseases [151]. Both of the above tests seemed to have good potential for rapid diagnosis of ALA and merit further multicentre validation using a much larger sample size.

Future perspectives

Quick and correct diagnosis of amoebic infection is important to avoid morbidity, death and disease transmission. Since most endemic areas are underdeveloped and lack resources (financial, facility, skilled manpower), tests which fulfilled the World Health Organization ASSURED criteria would be welcomed, i.e. affordable, highly sensitive and specific, user-friendly, robust and rapid, equipment-free and deliverable to those who need them [111]. Thus far, the tests developed for amoebiasis fulfilled some but not all of the above criteria. Thus, scientists working on laboratory diagnosis of amoebiasis should try to develop tests that fully comply with the above criteria.

Other than identifying infected individuals, it would be very useful to identify biomarkers that can predict individuals who are susceptible to acquire symptomatic infection, distinguish between the different stages of an infection, and monitor whether treatment leads to cure. The combination of well-characterized patient samples and advances in ‘omics’ technologies may make this possible. Genomics technologies may be able to assess host susceptibility based on the variations in single nucleotide polymorphism (SNP) of a certain loci [152]. Meanwhile, proteomics and metabolomics may help to distinguish the different stages of E. histolytica infection. Massive proteomic data on amoeba, especially E. histolytica, have been generated, whereby its proteome and sub-proteomes have been explored [153,154,155,156,157,158]. Hence, the biological processes of the E. histolytica proteins could be deduced. This will help us to better understand the role of certain E. histolytica proteins in the pathogenesis even though not all proteins are well characterized. Using quantitative proteomic, some E. histolytica proteins of a virulent variant have been confirmed to be differentially abundant when compared to its non-virulent variant or other non-pathogenic Entamoeba species [159, 160]. In addition, a comparison between E. histolytica life stages has been made using this approach [152]. Proteins that showed increased in abundance during certain life stages can be explored as potential biological markers for E. histolytica diagnosis, as well for drug and vaccine development.

Conclusion

Laboratory diagnosis of both intestinal and extraintestinal amoebiasis has improved with more sensitive and specific tests. For intestinal infection, molecular diagnostic is the way forward, with PCR-based assays for well-equipped laboratories and LAMP for low-resource settings. For extraintestinal infection, improvement in serodiagnosis is needed to accurately detect active infection since the exclusion of past infection is a challenge for the current IgG-based assays. Similarly, the sensitivity of rapid POC diagnostic tests for ALA can also be further improved and be made commercially available to enable its wider application. Further research areas that can be explored include discovery of biomarkers that can predict individuals prone to symptomatic infection, distinguish between the different stages of infection and monitor treatment success.