Comparison of Multispot EIA with Western blot for confirmatory serodiagnosis of HIV

Abstract

Background Recent improvements in the sensitivity of immunoassays (IA) used for HIV screening, coupled with increasing recognition of the importance of rapid point-of-care testing, have led to proposals to adjust the algorithm for serodiagnosis of HIV so that screening and confirmation can be performed using a dual or triple IA sequence that does not require Western blotting for confirmation. One IA that has been proposed as a second or confirmatory test is the Bio-Rad Multispot ® Rapid HIV-1/HIV-2 Test. This test would have the added advantage of differentiating between HIV-1 and HIV-2 antibodies. Objective To compare the sensitivity and type-specificity of an algorithm combining a 3rd generation enzyme immunoassay (EIA) followed by a confirmatory Multispot with the conventional algorithm that combines a 3rd generation EIA (Bio-Rad GS HIV-1/HIV-2 Plus O EIA) followed by confirmatory Western blot (Bio-Rad GS HIV-1 WB). Methods 8760 serum specimens submitted for HIV testing to the New York City Public Health Laboratory between May 22, 2007, and April 30, 2010, tested repeatedly positive on 3rd generation HIV-1-2+O EIA screening and received parallel confirmatory testing by WB and Multispot (MS). Results 8678/8760 (99.1%) specimens tested WB-positive; 82 (0.9%) tested WB-negative or indeterminate (IND). 8690/8760 specimens (99.2%) tested MS-positive, of which 14 (17.1%) had been classified as negative or IND by WB. Among the HIV-1 WB-positive specimens, MS classified 26 (0.29%) as HIV-2. Among the HIV-1 WB negative and IND, MS detected 12 HIV-2. Conclusion MS detected an additional 14 HIV-1 infections among WB negative or IND specimens, differentiated 26 HIV-1 WB positives as HIV-2, and detected 12 additional HIV-2 infections among WB negative/IND. A dual 3rd generation EIA algorithm incorporating MS had equivalent HIV-1 sensitivity to the 3rd generation EIA-WB algorithm and had the added advantage of detecting 12 HIV-2 specimens that were not HIV-1 WB cross-reactors. In this series an algorithm using EIA followed by MS would have resulted in the expedited referral of 38 specimens for HIV-2 testing and 40 specimens for nucleic acid confirmation. Further testing using a combined gold standard of nucleic acid detection and WB is needed to calculate specificity and validate the substitution of MS for WB in the diagnostic algorithm used by a large public health laboratory.