1. Introduction
Intestinal protozoa, including
Cryptosporidium spp.,
Entamoeba histolytica, and
Giardia duodenalis, contribute to the global burden of diarrhoeal diseases and are notably relevant to public health [
1,
2]. These pathogens are responsible for a wide range of gastrointestinal manifestations that lead to acute and chronic diarrhoea, abdominal pain, malnutrition, and diverse long-term sequelae [
3,
4,
5]. In addition, impaired linear growth and cognition have been documented in young children, particularly those living in resource-poor settings [
6,
7]. Both
Cryptosporidium spp. and
G. duodenalis are also major causes of waterborne and foodborne gastrointestinal disease outbreaks that are mostly detected in medium- to high-income countries [
8,
9]. Not surprisingly,
Cryptosporidium spp.,
G. duodenalis, and
E. histolytica are responsible for seven out of ten gastrointestinal parasites that are annually diagnosed in European clinical settings [
10]. Although
Entamoeba dispar is generally considered non-pathogenic, its common presence in human stool samples can, because of its morphological similarity to pathogenic
E. histolytica, impair diagnostic efforts [
11].
Blastocystis sp. and
Dientamoeba fragilis are other intestinal protists that have been recognized as potential contributors to intestinal and extraintestinal manifestations [
12,
13,
14], although the fact that both microorganisms are often found in individuals without obvious clinical manifestations complicates their consideration as primary pathogens.
Molecular assays based on real-time PCR (qPCR) are rapidly replacing conventional detection methods (e.g., microscopy) in the first line diagnosis of diarrhoea-causing intestinal protists that have a tendency to move rapidly from detection of a single pathogen to detection of multiple pathogens. This tendency is particularly evident in medium- to high-income countries with well-equipped clinical laboratories, where the gastrointestinal disease burden of protists is low and diagnostic sensitivity is an issue [
15,
16]. The benefits of qPCR include:(i) high throughput of stool screening, (ii) simultaneous detection of multiple pathogens within a single sample, and (iii) fast and accurate generation of results, enabling timely clinical decisions and swift interventions [
17,
18,
19,
20]. In resource allocation, qPCR assays minimize the need for repeated sample processing and analysis, reduce turnaround times, and improve workflows and operational efficiency in a cost-effective manner [
21,
22].
The validation of new commercially available diagnostic assays is one of the main tasks performed by national reference centres, which can bring together technical resources (e.g., biological samples for reference purposes, equipment) and expertise to perform the task efficiently. Here, we evaluated the clinical diagnostic performance of five singleplex and one duplex VIASURE Real-Time PCR assays in the detection and identification of common and clinically relevant intestinal protist parasites that affect the human gut, including Blastocystis sp., Cryptosporidium spp., Dientamoeba fragilis, Entamoeba dispar, Entamoeba histolytica, and Giardia duodenalis.
3. Results
Table 3 shows the agreement of results between the evaluated singleplex/duplex CerTest VIASURE Real-Time PCR detection kits and the reference PCR methods used in the initial diagnosis.
The singleplex VIASURE
Cryptosporidium spp. assay correctly identified 94.8% (91/96) of the DNA samples that tested positive for this pathogen (92 of human and 4 of wildlife origin, respectively). The assay recognized isolates belonging to six distinct
Cryptosporidium species, including primarily anthroponotic
C. hominis (
gp60 subtype families Ia, Ib, and Ie), zoonotic
C. parvum (
gp60 subtype families IIa, IIc, and IId), canine-adapted
C. canis, feline-adapted
C. felis, bovine-adapted
C. ryanae, and suine-adapted
C. scrofarum (
Table S1). It is noteworthy that 66.7% (64/96) of samples were monoinfected with
Cryptosporidium spp., and the remaining were concomitantly infected with
Blastocystis sp. (19.8%, 19/96), followed by
G. duodenalis (18.8%, 18/96),
D. fragilis (8.3%, 8/96) and
E. dispar (1.0%, 1/96) in nine different combinations (
Table S1).
Regarding
G. duodenalis, the singleplex VIASURE
Giardia assay accurately detected 96.5% (111/115) of the DNA samples positive for this protozoon, including zoonotic assemblages A and B, canine-adapted assemblage C, and feline-adapted assemblage F (
Table S1). Overall, 45.2% (52/115) of samples were monoinfected with
G. duodenalis and the remaining were concomitantly infected by
Blastocystis sp. (39.1%, 45/115), followed by
D. fragilis (29.6%, 34/115),
E. dispar (6.1%, 7/115), and
Cryptosporidium spp. (4.3%, 5/115), in eight different combinations (
Table S1).
Similarly, the singleplex VIASURE
E. histolytica assay correctly identified 96.0% (24/25) of the DNA samples positive for this pathogen (
Table S1). Overall, 76.0% (19/25) of samples were monoinfected with
E. histolytica. All six remaining
Entamoeba-positive samples were co-infected with
Blastocystis sp. (24.0%, 6/25) (
Table S1).
The singleplex VIASURE assays for the detection of
E. dispar and
D. fragilis correctly identified 100% (11/11 and 37/37, respectively) of the DNA samples positive for these protists (
Table S1). All 11
E. dispar-positive samples were co-infected with enteric protist species, including
Blastocystis sp. (90.9%, 10/11),
D. fragilis and
G. duodenalis (36.4%, 4/11 each) and
Cryptosporidium spp. (9.1%, 1/11), in five different combinations. Out of the 37
Dientamoeba-positive samples, 32.4% (12/37) corresponded to monoinfections by protozoan and the remaining were coinfected by
Blastocystis sp. (59.5%, 22/37),
G. duodenalis (10.8%, 4/37), and
E. dispar (8.1%, 3/37), in four different combinations (
Table S1).
With regard to the duplex VIASURE Real-Time PCR for the simultaneous detection of
D. fragilis and
Blastocystis sp., the assay correctly identified all (37/37 and 42/42, respectively) DNA samples positive for these protists (
Table S1). This assay performed equally well as the singleplex counterpart in detecting
D. fragilis. Of the 42
Blastocystis-positive samples, 81.0% (34/42) corresponded to monoinfections by this protozoan, and the remainder were coinfected by
G. duodenalis (16.7%, 7/42),
E. dispar (4.8%, 2/42), and
D. fragilis (2.4%, 1/42), in three different combinations (
Table S1).
The singleplex VIASURE assays yielded 10 false-negative results (five
Cryptosporidium spp., four
G. duodenalis, and one
E. histolytica) during testing. Successful EIC amplification of all 10 samples discarded the possibility of PCR inhibition. Reassessment of the five
Cryptosporidium spp. samples (four
C. hominis, one
C. parvum) with the reference PCR method yielded positive results in all five cases, which the VIASURE
Cryptosporidium assay confirmed to be false-negative results (
Table S1). Reassessments of the four
G. duodenalis samples and the single
E. histolytica sample with their corresponding reference PCR methods also yielded positive results (range of C
T values: 30.9–41.0) in all cases, and the VIASURE
Giardia and the VIASURE
E. histolytica assays confirmed they were false-negative results (
Table S1).
Of the 32 DNA samples used to determine potential cross-reactions, none yielded false-positive results when testing for
Cryptosporidium spp.,
G. duodenalis,
E. histolytica, and
E. dispar (
Table S1). However, four DNA samples cross-reacted with
Blastocystis sp., including isolates initially positive for
Cystoisospora sp. (
n = 1),
Leishmania infantum (
n = 1), and
Enterocytozoon bieneusi (
n = 2). An additional sample that was initially positive for
E. bieneusi cross-reacted with
D. fragilis (
Table S1). In addition, one human and one non-human primate sample harboured co-infections of
G. duodenalis and
Blastocystis, respectively, after both were previously detected by initial diagnosis (
Table S1).
Overall, very good agreement (Kappa test values ≥ 0.96) was observed between the results obtained by all VIASURE singleplex/duplex assays and those previously obtained by the reference PCR methods in the initial diagnosis (
Table 3).
Taking PCR results obtained during routine initial diagnosis as our reference, we summarize the diagnostic performance of the singleplex/duplex VIASURE assays evaluated here in
Table 4. In brief, sensitivity values for the five enteroparasites ranged from 0.94–1.00, specificity values from 0.97–1.00, positive predictive values from 0.89–1.00, and negative predictive values from 0.98–1.00.
Measures of diagnostic precision of repeatability (
Table 5) and reproducibility (
Table 6) were under acceptable ranges, with the exception of a positive
E. histolytica sample that failed to be amplified by the CerTest VIASURE
Cryptosporidium,
Giardia, &
E. histolytica Real-Time PCR detection kit.
4. Discussion
We carried out a comprehensive evaluation of the diagnostic performance of five singleplex (Cryptosporidium spp., D. fragilis, E. dispar, E. histolytica, and G. duodenalis) and a duplex (Blastocystis sp. and D. fragilis) CerTest VIASURE Real-Time PCR assays. Cryptosporidium spp., E. histolytica, and G. duodenalis are the three most clinically relevant intestinal protozoan parasites, and Blastocystis sp., D. fragilis, and E. dispar are common protist species of uncertain pathogenicity.
Most previous similar studies that evaluated commercial qPCR assays used prospectively collected stool samples that were submitted for routine investigation in clinical settings [
21,
23,
25,
26,
27,
28,
29,
30,
31,
32,
33]. In contrast, this survey purposely used a DNA panel with a wide diversity of parasite species and genotypes, with the aim of providing an evidence-based answer to the question of if the evaluated qPCR assays were suitable for the detection of less common or rare species/genotypes and animal-adapted genetic variants with known zoonotic potential and diverse geographical origins. This is important for intestinal protist species with large genetic diversities, such as
Cryptosporidium spp.,
G. duodenalis, and
Blastocystis sp.
Of 50 valid
Cryptosporidium species [
34], 19 have been reported in humans, with
C. hominis and
C. parvum accounting for ca. 90% of the human cases of cryptosporidiosis reported globally [
35]. The VIASURE
Cryptosporidium assay was able to detect
C. hominis (including subtype families Ia, Ib, and Ie),
C. parvum (including subtype families IIa, IIc, and IId), and animal-adapted
C. canis,
C. felis,
C. ryanae, and
C. scrofarum. The diagnostic sensitivity and specificity of this assay in the detection of
Cryptosporidium spp. were 0.94 and 1, respectively. These values were very similar to those obtained with the singleplex and multiplex versions of the assay in previous studies (0.96–1 and 0.99–1, respectively) [
23,
32]. The sensitivity performance of the VIASURE
Cryptosporidium assay is also comparable with those (0.96–1) documented for other commercially available qPCR kits, including the EasyScreen Enteric Parasite Detection Kit (Genetic Signatures, Sydney, Australia) and the FilmArray Gastrointestinal Panel (BioFire Diagnostics, Salt Lake City, UT, USA) [
21,
27]. Lower scores (0.53–0.96) were obtained with the BD MAX Enteric Parasite Panel (Becton, Dickinson and Company, Franklin Lakes, NJ, USA), the FTD Stool Parasites (FAST-Track Diagnostics, Esch-sur-Alzette, Luxembourg), the Gastroenteritis/Parasite Panel I (DIAGENODE, Liège, Belgium), and the RIDA
®GENE Parasitic Stool Panel II (R-Biopharm AG, Pfungstadt, Germany), assays, respectively [
10,
25,
26,
31,
32].
Giardia duodenalis consists of eight distinct assemblages (A–H) that differ in host specificity, with assemblages A and B being most commonly reported in humans and several other mammal species [
36,
37]. Sporadic cases of human infection, by canine-adapted assemblages C/D, feline-adapted assemblage F and ungulate-adapted assemblage E, have also been reported, particularly in children and immunocompromised patients [
38,
39]. The VIASURE
Giardia assay was able to identify most of the aforementioned
G. duodenalis assemblages (no D and E isolates were available for testing in this study), demonstrating its usefulness in detecting human and zoonotic genetic variants of the parasite. The diagnostic sensitivity and specificity of The VIASURE
Giardia assay were 0.94 and 0.99, respectively. More variable results (sensitivity: 0.81–0.97; specificity: 0.94–1) were obtained with the singleplex and multiplex versions of the assay by previous studies [
23,
32]. It is important to note that other commercially available qPCR assays for the detection of
G. duodenalis have consistently achieved diagnostic sensitivities over 0.97, including the BD MAX [
25,
26], the FilmArray [
21], and the NanoCHIP Gastrointestinal Panel (Savyon Diagnostics, Ashdod, IL, USA) [
29,
30] methods. In contrast, the Gastroenteritis/Parasite Panel I has shown poorer performance (0.68–0.76) [
31,
32].
Entamoeba histolytica displays a relatively low level of nucleotide diversity in non-repetitive loci and is not subdivided into intra-species genetic variants for genotyping purposes [
40]. This limited genetic variability eases the molecular detection of the parasite. The VIASURE
E. histolytica assay achieved sensitivity and specificity values of 0.96 and 1, respectively, which is consistent with previous studies, which obtained similar values through the same assay and its multiplex variant (sensitivity: 0.96–1; specificity: 1) [
23,
32].
Similar, but slightly lower, diagnostic sensitivities (0.92–0.95) have also been reported after the BD MAX [
25,
26], EasyScreen [
21], and Luminex xTAG Gastrointestinal Pathogen Panel (Luminex Molecular Diagnostics, Toronto, ON, Canada) [
29] methods were used.
E. dispar, which is morphologically identical to
E. histolytica and closely related to it, is a common finding in human stools that is thought to be avirulent [
41], meaning differential diagnosis of
E. histolytica and
E. dispar is important in areas where both species occur sympatrically. However, few commercial qPCR-based methods are available for this purpose [
32,
42], and the VIASURE
E. dispar assay achieved sensitivity and specificity values of 1, higher than those (sensitivity: 0.95–0.96; specificity: 1) obtained by similar previous studies [
32,
42].
Dientamoeba fragilis isolates also have a very low level of genetic variability, regardless of geographic area of origin or the presence/absence of clinical manifestations. To date, two
D. fragilis genotypes (1 and 2) have been recognized, with a strong predominance of genotype 1, in humans [
43]. The VIASURE
D. fragilis assay achieved sensitivity and specificity values of 1 and 0.99 (respectively) in both singleplex and duplex versions. This diagnostic performance was in the higher range of what was previously documented in the use of other commercially available qPCR assays (sensitivity: 0.90–0.96; specificity: 1) [
32,
33].
Blastocystis sp. is a frequent protist that colonizes/infects the human gastrointestinal tract. To date, 40
Blastocystis subtypes, including ST1-ST17, ST21, and ST23-ST44, are considered taxonomically valid [
44,
45,
46,
47,
48]. The VIASURE
Blastocystis +
D. fragilis assay was able to identify ST1–ST4 (the most frequent STs circulating in humans), in addition to ST5 and ST8. The diagnostic sensitivity and specificity of The VIASURE
Giardia assay were 0.94 and 0.99, respectively. Overall, these figures were superior to those obtained with other commercial qPCR assays, including the Allplex Gastrointestinal Panel-Parasite assay (Seegene, Seoul, Korea; sensitivity: 0.84–1; specificity: 0.81–082) [
49,
50] or the LIGHTMIX Gastro Parasite assay (Tib MolBiol, Berlin, Germany; sensitivity: 0.99; specificity: 0.89) [
49]. Here it should be noted that the diagnostic performance of the latter assay did not consider the subtype of the isolates investigated.
Direct comparison of results obtained in different studies that evaluate the diagnostic performance of commercial qPCR assays should consider a number of factors, including panel sample size (low sample numbers are likely to result in inaccurate and inconsistent estimates) and composition (genetic diversity of rare or less frequent species/genotypes might affect amplification success rates), sample (reflecting parasite load and sometimes virulence/pathogenicity), and the diagnostic method used as a gold standard.
Part of this study’s advantage over predecessors lies in its careful selection of a large panel of molecularly confirmed DNA samples (PCR and Sanger sequencing) for analysis. This advantage notwithstanding, we are aware that some relevant pathogenic and commensal protozoan species were missing from our panel, including C. meleagridis (the third most common cause of cryptosporidiosis in humans) and also potentially cross-reactive species, including Cyclospora cayetanensis, Entamoeba coli, Endolimax nana and Encephalitozoon intestinalis, amongst others. Future studies should address this issue of missing species.