ELISA absorbance cut-off method affects malaria sporozoite rate determination in wild Afrotropical Anopheles

Malaria sporozoite infection rates in a mixed species group of 244 Anopheles gambiae Giles sensu lato and 115 An.funestus Giles wild female mosquitoes were compared using three methods to determine cut-off absorbance values for positivity of a Plasmodium falciparum Welch enzyme-linked immunosorbent assay (ELISA). Positive controls were based on P.falciparum circumsporozoite protein. As negative controls, four wild male Anopheles were included on each microtitre plate; tests were repeated on four consecutive days for each plate. Infection rates were estimated at 13.1-22.8% using the mean absorbance value of negative controls plus three standard deviations, 11.7-12.8% using double the mean and 12.5-13.6% using the fixed cut-off value of 0.20 (allowing for 20% variation in negative control absorbance values). Observed agreement for positivity or negativity among samples tested four times was 98.6% for the 2 x mean method, 97.2% for the fixed cut-off 0.20 value, but only 82.7% for the mean + 3 SD method. It was concluded that the 2x mean cut-off method is most reliable for field studies. P.falciparum sporozoite rates of 12.2% in An.funestus and 11.9% in An.gambiae s.l. were thus determined on the basis of the 2x mean cut-off method. This comparative evaluation demonstrates that vector infectivity rates can be seriously over-estimated from sporozoite ELISA tests, by as much as 87% in one case considered here, depending on the absorbance cut-off method applied for negative controls.