The performance of antibody detection for the diagnosis of schistosomiasis has been evaluated in Kenya. Approximately 1500 blood samples from 3 areas with endemic schistosomiasis ( Schistosoma mansoni only, S. haematobium only, and a mixed infection area), and from a non-endemic control area, were tested for their antibody reactivity in an enzyme-linked immunosorbent assay (ELISA). The results were compared with infection status determined by parasitological examination. Two test antigens were used: unfractionated S. mansoni egg homogenate (SEA), and CEF6, a previously described, partially purified fraction of SEA containing 2 cationic antigens. The antigens prepared from eggs of Kenya and Puerto Rico S. mansoni isolates gave very similar results. Bloods from patients with S. haematobium infection cross-reacted significantly with the two S. mansoni antigen preparations, but reactivity against CEF6 appeared more specifically indicative of S. mansoni infection. Of 254 blood samples from schoolchildren in the non-endemic area, 100% gave ELISA optical density readings at 492 nm (OD 492) <0·20 against SEA, and 98% were <0·20 against CEF6. With 887 blood samples from subjects of all ages in the area endemic for S. mansoni alone, using an ELISA OD 492 cut-off point of 0·20, SEA and CEF6 had sensitivities of 94% and 97% respectively, and specificities of 64% and 59% respectively. Increasing the OD 492 cut-off value reduced the sensitivity and increased the specificity of both test antigens. Specificity of both antigens was poor with samples from 234 children in an area endemic for both S. mansoni and S. haematobium (<20% for both antigens at an OD 492 cut-off value of 0·20). It is suggested, with supporting evidence, that one reason for the apparently poor specificity of the serological testing in the endemic area is the inherently poor sensitivity of parasitological examinations of small volumes of stool. There was a significant positive correlation between blood elisa results and the number of eggs excreted by infected subjects in the area endemic for S. mansoni only. Highest correlation coefficients were obtained in children aged under 10 years, and CEF6 gave marginally higher correlation coefficients than SEA. The graphs of prevalence and intensity of schistosome infection drawn from serological results were similar in shape to the graphs of these 2 quantities based on parasitological results, and the results indicate that serology merits wider use as an epidemiological tool for determining infection status in schistosomiasis.
Read full abstract