Beier J C, Perkins P V, Koros J K, Onyango F K, Gargan T P, Wirtz R A, Koech D K, Roberts C R
Kenya Medical Research Institute, Nairobi.
J Med Entomol. 1990 May;27(3):377-84. doi: 10.1093/jmedent/27.3.377.
Malaria infection rates determined by dissection and Plasmodium falciparum enzyme-linked immunosorbent assay (ELISA) were compared for 26,935 Anopheles gambiae Giles sensu lato and 17,739 Anopheles funestus Giles collected during 20 mo in western Kenya. ELISA infection rates were about 43% higher than dissection sporozoite rates. In dissection-negative Anopheles, circumsporozoite (CS) protein was detected by ELISA in 5.2% of 10,017 salivary gland samples and in 12.2% of 237 thorax samples. The accuracy of dissection and ELISA techniques was compared by the following tests on a group of 352 field-collected Anopheles (held 10 d to ensure sporogonic development): salivary gland dissection, examination of Giemsa-stained dissection slides, ELISA tests on salivary gland and thorax body parts, and microscopic techniques for determining sporozoite loads. Respective infection rates were 9.9%, 10.8%, and 15.6% for dissection, stained slides, and ELISA. Sporozoite loads were associated significantly with ELISA absorbance values (r = 0.76). Compared with Giemsa-stained dissection slide results, the sensitivity of sporozoite detection was 92.1% for dissection compared with 78.9% for ELISA; specificity was 100.0% for dissection versus 92.0% for ELISA. Immunological detection of CS protein in head-thorax samples of Afrotropical vectors overestimated the proportion of infective Anopheles because the comparison of techniques indicated that 45.4% of the ELISA positive Anopheles did not contain salivary gland sporozoites.
