Application of a recombinase polymerase amplification (RPA) assay and pilot field testing for Giardia duodenalis at Lake Albert, Uganda

BACKGROUND: Giardia duodenalis is a gastrointestinal protozoan causing 184 million cases of giardiasis worldwide annually. Detection is by microscopy or coproantigen assays, although sensitivity is often compromised by intermittent shedding of cysts or trophozoites, or operator expertise. Therefore, for enhanced surveillance field-applicable, point-of-care (POC), molecular assays are needed. Our aims were to: (i) optimise the recombinase polymerase amplification (RPA) assay for the isothermal amplification of the G. duodenalis β-giardin gene from trophozoites and cysts, using published primer and probes; and (ii) perform a pilot field validation of RPA at a field station in a resource-poor setting, on DNA extracted from stool samples from schoolchildren in villages around Lake Albert, Uganda. Results were compared to an established laboratory small subunit ribosomal RNA (SSU rDNA) qPCR assay with additional testing using a qPCR targeting the triose phosphate isomerase (tpi) DNA regions that can distinguish G. duodenalis of two different assemblages (A and B), which are human-specific. RESULTS: Initial optimisation resulted in the successful amplification of predicted RPA products from G. duodenalis-purified gDNA, producing a double-labelled amplicon detected using lateral flow strips. In the field setting, of 129 stool samples, 49 (37.9%) were positive using the Giardia/Cryptosporidium QuikChek coproantigen test; however, the RPA assay when conducted in the field was positive for a single stool sample. Subsequent molecular screening in the laboratory on a subset (n = 73) of the samples demonstrated better results with 21 (28.8%) RPA positive. The SSU rDNA qPCR assay resulted in 30/129 (23.3%) positive samples; 18 out of 73 (24.7%) were assemblage typed (9 assemblage A; 5 assemblage B; and 4 mixed A+B). Compared with the SSU rDNA qPCR, QuikChek was more sensitive than RPA (85.7 vs 61.9%), but with similar specificities (80.8 vs 84.6%). In comparison to QuikChek, RPA had 46.4% sensitivity and 82.2% specificity. CONCLUSIONS: To the best of our knowledge, this is the first in-field and comparative laboratory validation of RPA for giardiasis in low resource settings. Further refinement and technology transfer, specifically in relation to stool sample preparation, will be needed to implement this assay in the field, which could assist better detection of asymptomatic Giardia infections. [Image: see text]