Supplementary anti-hepatitis C virus (HCV) testing with 2nd and 3rd generation recombinant immunoblot assay and matrix applied to enzyme immunoassay positive sera and comparison with HCV-RNA detection.

A series of 118 serum samples tested positive in the anti-hepatitis C virus (HCV) 2nd-generation enzyme immunoassay (EIA-2), and indeterminate (93 samples) or negative (25 samples) in the supplementary 2nd-generation recombinant immunoblot assay (RIBA-2). These sera were further evaluated by three additional tests: 3rd-generation RIBA (RIBA-3), MATRIX, and AMPLICOR HCV-PCR. For the 93 RIBA-2 indeterminate serum samples, the results of the immunoassays had a concordance of 69%. Twenty-one and 34 samples remained anti-HCV indeterminate in the RIBA-3 test and the MATRIX, respectively. Among the 25 RIBA-2-negative samples, only seven samples remained anti-HCV negative, while five samples tested anti-HCV positive in both RIBA-3 and MATRIX. The reactivity of the RIBA-3 antigen NS5 was not crucial for the result of any sample. Positive to negative contradictions between the results of MATRIX and RIBA-3 were never observed. Altogether, the MATRIX tested a significantly lower number of samples anti-HCV negative than did the RIBA-3. HCV RNA was detectable in 54/93 RIBA-2 indeterminate and 7/25 RIBA-2 negative samples. High percentages of PCR positive results among RIBA-3-indeterminate and among MATRIX-indeterminate samples indicate an increased possibility of detecting HCV RNA if at least one antigen is reactive. The type of antigen, the pattern of antigen reactivity, or the level of reactivity had no prognostic value in predicting the presence of HCV RNA. Our findings show the necessity of being cautious in the interpretation of RIBA-2-negative results.