Because HCV infection was associated mostly with Gammagard produced solely from plasma screened by EIA-2, we investigated whether such screening might have removed antibodies to HCV envelope proteins, E1 and E2, from plasma and the resulting IGIV lots. Recombinant baculovirus constructs derived from the HCV H-strain (genotype 1a), HCV-Bac 3 and HCV-Bac 5, were used to express E1 and E2 proteins in SF-9 insect cells. The expressed proteins were partially purified by GNA-Lectin affinity chromatography and characterized by Western blots with monoclonal antibodies against E1 and E2. Both isolated E1 and E2 proteins were N-glycosylated. HCV-Bac 3 expressed E1 as a fusion protein (E1-PH) while HCV-Bac 5 expressed a full-length E1 and E2 fusion protein (E2-PH). Both E1-PH and E2-PH were immunoreactive with human plasma and IGIV lots tested while the full-length E1 was not. Thus proteins derived from HCV-Bac 3 and HCV 5 were subsequently used to establish an ELISA for detection of anti-E1 and anti-E2, respectively. Both antibodies were present in a pool of 186 EIA-1 reactive plasma units, but levels were moderate compared with those in plasma from a patient with chronic HCV infection. Very low levels of these antibodies were detected in a pool consisting of 2,887 EIA-1 screened or 7,000 EIA-2 screened units. Of 180 plasma samples which were EIA-1 reactive, 63% were positive for anti-E1, 73% for anti-E2, and 58% for both. However, 6% of the anti-E1 positive samples were devoid of anti-E2 while 16% of anti-E2 positive samples were negative for anti-E1. Among IGIV lots assayed, highest titers of both antibodies were detected in 2 experimental hepatitis C IGIV (HCIGIV) lots which include one prepared by North American Biologicals from 198 EIA-2 reactive units and another prepared in our laboratory from 186 EIA-1 reactive units. Gammagard lots manufactured from unscreened plasma contained lower levels of both antibodies than either HCIG lot but had much higher titers than those from either EIA-1 or EIA-2 screened plasma. No significant differences were observed in plasma or in IGIV lots prepared from either type of screened plasma. Our data suggest that screening greatly reduced levels of antibodies to both E1 and E2 in plasma and the resulting IGIV. The lack of these complexing and/or neutralizing antibodies may have contributed to the adverse outcome of Gammagard.
