Human Immunodeficiency Virus (HIV) Antigen-Antibody Combination Assays: Evaluation of HIV Seroconversion Sensitivity and Subtype Detection

Recently, 4th generation assays that permit the combined detection of human immunodeficiency virus type 1 (HIV-1) p24 antigen (Ag) and anti-HIV antibody (Ab) were introduced on the international market. The diagnostic window is reduced on average by 4 days with this new assay generation in comparison to 3rd generation antibody-screening enzyme immunoassays (EIAs) in HIV-1 group M subtype B primary infections (1, 6). Up to now, no data had been available on the sensitivity of this new generation of assays for detection of HIV-1 non-B subtype primary infection.

Dilution series of virus lysates of HIV-1 group M non-B subtypes, group O, and HIV-2, in HIV-negative serum are used as a substitute for seroconversion panels for the assessment of the influence of genetic diversity on the sensitivity of 4th generation HIV assays. This seems a valuable approach, since the sensitivity of 4th generation assays in seroconversion panels correlates with sensitivity for Ag detection in virus lysates at least for HIV-1 subtype E primary infection (unpublished data). However, the results from dilutions of virus lysates must be interpreted with caution if a limited number of virus stocks are tested.

The sensitivities for HIV Ag detection of one prototype and six commercially available HIV Ag-Ab (4th generation) combination assays were evaluated by testing virus stocks (n = 10) of different HIV-1 group M subtypes, group O, and HIV-2 diluted in HIV-negative serum adjusted to concentrations ranging between 5 and 15 pg of p24 Ag per ml according to the Dupont standard (Table 1). The sensitivities of the 4th generation assays were highly variable. The best performance was obtained with the VIDAS HIV DUO Ultra and Murex HIV Ag-Ab combination, which detected 7 of 10 virus lysates. Four assays were not able to detect HIV Ag in any 1 of the 10 isolates. The failure of VIDAS DUO HIV Ultra to detect HIV-1 subtype C Ag under a concentration of 34 pg of p24 Ag per ml (which corresponds appreciatively to 125 pg of HIV Ag per ml) as observed by Ly et al. (2) could not be confirmed by our investigation. In a previous study from our group (6), VIDAS HIV DUO Ultra also showed a high sensitivity for Ag detection in virus isolates of different subtypes (A to H and group O). Its sensitivity for two HIV-1 subtype C isolates was equivalent to that of a stand-alone Ag assay with a detection threshold of 3 pg of p24 Ag per ml. The results of dilutions of HIV-1 subtype virus lysates are dependent on the strain used and cannot be generalized for all isolates of an HIV-1 subtype. In this context, it should be underlined that VIDAS HIV DUO Ultra failed to detect one HIV-1 group O isolate at a concentration of 6 pg of p24 Ag per ml (Table 1). However, in a previous evaluation, the assay showed a more than fourfold-higher sensitivity than the comparative stand-alone Ag assay for two of three group O isolates. Conclusions on the sensitivity of Ag detection from different HIV-1 subtypes can only be drawn by testing larger numbers of virus stocks than only one isolate of each subtype.

TABLE 1.

End point titration of HIV-infected cell culture supernatants

Culture supernatant	Titer of p24 Ag in pg/ml (pg of HIV Ag/ml)a	Sub- or serotype	Test result, expressed as index (signal/cutoff value)b
			VIDAS HIV DUO Ultra	Murex HIV Ag-Ab combination	Cobas Core HIV Combi	VIDAS HIV DUO	Genscreen Plus HIV Ag-Ab	Enzygnost Integral	Vironostika HIV Uniform II Ag-Ab	Cobas Core HIV Ag EIA	Elecsys HIV Ag	Coulter HIV-1 p24 Ag
MP33	8 (29)	A	2.28	0.96	1.14	0.76	0.34	0.12	0.55	2.74	2.09	2.48
MP77	5 (18)	B	1.12	0.60	0.61	0.60	0.41	0.10	0.53	1.28	1.34	1.59
MP37	6 (22)	C	1.96	1.16	0.52	0.48	0.31	0.05	0.54	1.50	1.65	1.49
VI824	11 (40)	D	3.04	1.20	1.41	0.80	0.39	0.17	0.49	3.25	2.76	2.92
MP48	5 (18)	E (CRF A/E)	1.24	0.52	0.67	0.56	0.32	0.10	0.55	1.39	1.47	0.58
VI1310	9 (33)	F	1.24	1.19	1.14	0.60	0.44	0.31	0.54	2.59	2.24	1.81
VI1197	8 (29)	G	1.72	1.20	0.95	0.60	0.28	0.28	0.56	2.19	1.93	0.66
VI991	6 (22)	H	0.2	1.98	1.23	0.32	0.34	0.13	0.50	2.97	1.53	1.84
MP331	6 (22)	O	0.68	1.16	0.88	0.40	0.29	0.10	0.52	2.12	1.65	0.47
VI390	15 (54)	HIV-2	0.36	1.65	0.44	0.24	0.15	0.06	0.47	1.80	3.06	0.12

^aCalibrated with Dupont standard material (1 pg of p24 Ag corresponds to 3.6 pg of total HIV Ag).

^bA value of ≥1 is considered positive.

Ly et al. tested the sensitivity of Ab detection to HIV variants by using dilution series of anti-HIV-1- and anti-HIV-2-positive serum samples. While dilution series of virus lysates are appropriate to challenge the sensitivity of the Ag detection module of 4th generation assays, the results obtained from dilutions of anti-HIV Ab-positive samples very probably do not reflect the sensitivity of the Ab detection module of these assays. The results correlate with the absolute number of HIV Ab molecules present in the serum; however, the method, which detects the lowest concentration of HIV Abs, may not necessarily show the highest sensitivity for seroconversion samples (4; B. J. Weiblen, J. Acker, P. E. Garrett, R. T. Schumacher, supplemental technical bulletin, Boston Biomedica, Inc., Boston, Mass., 1993). Assays that use large recombinant Ags with multiple Ab binding sites show a higher sensitivity than those using short synthetic peptides. For screening purposes, it is most important for assays to have the antigenic epitopes most strongly recognized earliest in seroconversion and persistently recognized throughout infection. Endpoint titration of anti-HIV-positive samples should not be used to evaluate test kit sensitivity.