Research Article: Comparative performance evaluation of four commercial multiplex real-time PCR assays for the detection of the diarrhoea-causing protozoa Cryptosporidium hominis/parvum, Giardia duodenalis and Entamoeba histolytica

Date Published: April 8, 2019

Publisher: Public Library of Science

Author(s): Silvia Paulos, José María Saugar, Aida de Lucio, Isabel Fuentes, María Mateo, David Carmena, Rodney D. Adam.


Multiplex molecular panels are relentlessly replacing conventional methods for the detection of enteric pathogens from stool samples in clinical and research laboratories. Here we evaluated four commercial multiplex real-time PCR assays for the detection of Cryptosporidium hominis/parvum, Giardia duodenalis and Entamoeba histolytica.

The diagnostic performance of the Gastroenteritis/Parasite Panel I (Diagenode), the RIDAGENE Parasitic Stool Panel (R-Biopharm), the Allplex Gastrointestinal Parasite Panel 4 (Seegene) and the FTD Stool Parasites (Fast Track) real-time PCR methods was assessed against a reference panel of 126 well-characterized DNA samples including Cryptosporidium hominis (n = 29), Cryptosporidium parvum (n = 3), Giardia duodenalis (n = 47), Entamoeba histolytica (n = 3), other parasite species (n = 20), and apparently healthy subjects (n = 24).

Obtained diagnostic sensitivities ranged from 53–88% for Cryptosporidium hominis/parvum, and from 68–100% for G. duodenalis. The R-Biopharm method achieved the best performance for the detection of Cryptosporidium hominis/parvum both in terms of diagnostic sensitivity (87.5%) and detection limit (a 100-fold increase compared to other tests). The Fast Track method was particularly suited for the detection of G. duodenalis, achieving a 100% sensitivity and a detection limit at least 10-fold superior. Detection of E. histolytica was similarly achieved by all compared methods except Diagenode.

Diagnostic performance varied largely depending on the method used and the targeted pathogen species. Factors including test sensitivity/specificity, cost, patient population surveyed, laboratory workflow, and diagnostic algorithm should be carefully considered when choosing the most appropriate multiplex PCR platform.

Partial Text

Enteric Giardia duodenalis, Cryptosporidium spp., and Entamoeba histolytica are the most important diarrhoea-causing protozoa globally. Infections by these parasites cause significant morbidity and mortality primarily among children living in resource-poor settings in developing countries [1], but are also a significant public health concern in developed nations [2]. Indeed, these three protozoan species account for up to 70% of the gastrointestinal parasites diagnosed every year at hospital-based microbiology laboratories in Europe [3]. Additionally, both G. duodenalis and Cryptosporidium spp. are increasingly recognized as important waterborne and foodborne pathogens all over the world [4–6].

The diagnostic performance results of the four multiplex qPCR methods compared here are summarized in Table 2. The R-Biopharm method was the most sensitive (87.5%) assay for the detection of Cryptosporidium hominis/parvum, with the Fast Track assay performing poorly (53.1%). All four methods detected the three C. parvum DNA samples assessed. When tested undiluted, Cryptosporidium-positive DNA samples generated an elevated number of inhibitory reactions mostly resolved when re-tested in a 1:10 dilution. This was particularly true for the R-Biopharm and the Diagenode methods (of note, the former manufacturer specifically recommends diluting faecal suspensions 1:3 prior to DNA extraction). This issue was observed neither for Giardia- nor Entamoeba-positive DNA samples. One-way ANOVA test results showed significant (P < 0.05) variations in the distribution of obtained Ct values among the four multiplex qPCR methods assessed here, although no one-by-one direct comparison of methods was attempted due to unsurmountable differences in equipment (e.g. thermocycler used) features and assay (e.g. initial volume of DNA tested) procedures. In line with the findings mentioned above, the R-Biopharm method produced lower mean Ct values either in unpaired (mean: 30.1; 95% CI: 28.6–31.5) (Fig 1A) and paired (mean: 28.2; 95% CI: 26.9–29.6) (Fig 1B) Cryptosporidium-positive DNA samples than the other three methods tested. In this comparative study we evaluated four commercial multiplex qPCR methods for the identification of the three most clinically relevant protozoan enteric parasites, namely Cryptosporidium spp., Giardia duodenalis, and Entamoeba histolytica. The diagnostic performance of both the Diagenode and the Fast Track assays have been individually assessed and compared to that obtained by microscopy in previous studies [3,13,15]. However, this is the first comprehensive survey reporting the diagnostic performance of the R-Biopharm and Seegene methods. Another significant methodological contribution of this work is that, in an attempt to normalize starting experimental conditions and strengthen the robustness of the obtained data, analyses were conducted against a reference panel of well-characterize DNA samples. Our results evidenced marked detection differences among tests depending on the targeted parasite species considered, with diagnostic sensitivities typically ranging from 53–88% for Cryptosporidium spp., and from 68–100% for G. duodenalis. The diagnostic sensitivity of these methods for the detection of E. histolytica was not fully assessed due to the insufficient number of positive samples to this pathogen available. Despite the highly variable sensitivity values observed, all four methods did not cross-react with any of the DNA samples from other protozoan or helminthic parasite species tested, suggesting a specificity near 100%. However, the exact extent of this statement should be confirmed in future studies including a larger panel of DNA samples from other potentially cross-reacting enteric pathogen and commensal species including Cyclospora cayetanensis, Entamoeba coli, Endolimax nana, Blastocystis sp., and Dientamoeba fragilis.   Source:


Leave a Reply

Your email address will not be published.