Comparison of commercial latex agglutination and sandwich enzyme immunoassays with a competitive binding inhibition enzyme immunoassay for detection of antigenemia and antigenuria in a rabbit model of invasive aspergillosis

A commercial latex agglutination assay (LA) and a sandwich enzyme immunoass ay (SEIA) (Sanofi Diagnostics Pasteur, Marnes-la-Coquette, France) were com pared with a competitive binding inhibition assay (enzyme immunoassay [EIA] ) to determine the potential uses and limitations of these antigen detectio n tests for the sensitive, specific, and rapid diagnosis of invasive asperg illosis (IA). Toward this end, well-characterized serum and urine specimens were obtained by using a rabbit model of IA, Serially collected serum or u rine specimens were obtained daily from control rabbits or from rabbits imm unosuppressed and infected systemically with Aspergillus fumigatus. By 4 da ys after infection, EW, LA, and SEIA detected antigen in the sera of 93, 93 , and 100% of A. fumigatus-infected rabbits, respectively, whereas antigen was detected in the urine of 93, 100, and 100% of the rabbits, respectively , False-positive results for non-ii, fumigatus-infected rabbits for EIA, LA , and SEIA were as follows: for serum,14, 11, and 23%, respectively; for ur ine, 14, 84, and 90%, respectively. Therefore, although the sensitivities o f all three tests were similar, the specificity was generally greater for E W than for LA or SEIA, infection was also detected earlier by EW, by which the serum of 53% of A. fumigatus-infected rabbits was positive as early as 1 day after infection, whereas the serum of only 27% of the rabbits tested by LA was positive. Although the serum of 92% of A. fumigatus infected rabb its was positive by SEIA as early as 1 day after infection, the serum of a high percentage (50%) was false positive before infection. The urine of 21% of A. fumigatus-infected rabbits was positive by EIA as early as I day aft er infection, and the urine of none of the rabbits was false positive befor e infection. When EIA results for urine specimens were combined with those for serum, sensitivity was improved (i.e., 67% of rabbits were positive by I day after infection and only one rabbit gave a false-positive result). A total of 93% of A. fumigatus-infected rabbits were positive for antigen In urine as early as 1 day after infection and the urine of 100% of the rabbit s was positive by SEIA, However, before infection, 79% of A. fumigatus-infe cted rabbits were false positive for antigen in urine by LA and 90% were fa lse positive for antigen in urine by SEIA. These data indicate that the EIA has the potential to be used to diagnose IA with both serum and urine spec imens and to detect a greater number of infections earlier with greater spe cificity than the specificities achieved with the commercial tests.