EVALUATION OF VERO CELL LYSATE ANTIGEN FOR THE ELISA OF FLAVIVIRUSES

The vero cell lysate antigen for the enzyme-linked immunosorbent assay (ELISA) of flaviviruses was evaluated for sensitivity, specificity in cluding cross-reactions, and background by comparing with the standard ELISA. Human sera, in serial dilutions, were taken from subjects 14, 35, and 210 days postvaccination with 17D antigen. Early after injecti on, high sensitivity (82.9%) was shown by the cell lysate antigen meth od. Late after infection, high sensitivity was achieved by the standar d method (96.2% and 94%), with significant difference (P = 0.0001). Ho wever, sensitivity achieved by the cell lysate antigen method was also acceptable (91.7% & 88.9%). The cell lysate antigen method showed hig h specificity and low cross reactivity early after infection. At 35 da ys postvaccination, no difference in specificity was observed between the two methods, but higher cross-reactions were observed for the stan dard method. This pattern continued at 210 days postvaccination, with significantly higher cross-reactions with the standard ELISA. The opti cal density differences by the two methods did not show significant re lationship with the serial dilutions of human sera. No difference was observed in early and late infections in the background values of the negative control (Western equine encephalitis) between the two methods . The ELISA by the cell lysate antigen, within the limits of the exper iments done, was found to be a good replacement for the ELISA by the s tandard method. (C) 1993 Wiley-Liss, Inc.