The use of genotyping assays for the detection and evaluation of drug resistance mutations within the polymerase gene of human immunodeficiency virus type 1 (HIV-1) is becoming increasingly relevant in the clinical management of HIV-1 infection. However, genotypic resistance assays available currently have been optimised for genetic subtype B strains of the virus and many clinical centres are presented with strains from subtypes A, C, and D. In the present report, we compare the performance of two sequence-based commercially available kits, the ViroSeq Genotyping System (Applied Biosystems, Foster City, CA) and the TruGene HIV-1 Genotyping Kit (Visible Genetics, Toronto, Ontario) against a panel of 35 virus isolates from HIV-1 Group M (subtypes A-J). Full-length consensus sequences were generated by the ViroSeq genotyping system for 26 of 31 (83.8%) of the isolates tested, in contrast to the TruGene genotyping system, which generated 16 of 30 (53%) usable sequences overall. Overall, subtype B isolates were sequenced with a greater degree of success than non-subtype B isolates. Discrepancies were found between the consensus sequences reported by each system for each sample (mean difference 1.0%; range 0.0-3.2%), but these appeared to be random and did not affect interpretation of the major resistance codons. In addition, both systems were able to amplify template RNA from low copy viral load plasma samples (10(2)-10(3) RNA copies/ml) taken from a random selection of patient samples encompassing subtypes A-C. While the availability of these genotyping systems should facilitate studies of HIV-1 drug resistance in countries in which these subtypes are prevalent, the performance against subtypes other than B needs to be improved.

HIV-1 is characterised by extensive genetic variability encompassing at least 10 different phylogenetically related clades within the major group of HIV-1 subtypes. Most commercially available HIV-1 RNA plasma viral load assays have been optimised with clade B viruses and may yield misleadingly low RNA levels for nonclade B viruses that are increasingly found in Europe. In this study we compare the most recent versions of the Roche Amplicor HIV Monitor and the Chiron Quantiplex for ability to detect viraemia in a population of patients infected with a range of HIV-1 subtypes. EDTA-treated plasma was obtained from 206 patients. The Amplicor and Quantiplex assays were carried out in accordance with manufacturers' instructions. Results from 53/206 (25.7%) samples differed by >0.4 log between Amplicor 1.5 and Quantiplex 3.0. A >0.5 log and 1.0 log difference was detected between Amplicor 1.5 and Quantiplex 3.0 in 37/206 (17.9%) and 7/206 (3.4%) of samples, respectively. Overall, Amplicor 1.5 gave a median value of 0.22 log higher than Quantiplex 3.0. Discordant results were detected in 53 out of 206 (25.7%) samples. Of these 22 out of 123 (17.9%) samples were of UK origin, 18 out of 43 (41.9%) African, 1 out of 8 (12.5%) South American, 1 out of 6 (16.7%) North American, 4 out of 9 (44.4%) North European, 3 out of 11 (23.7%) South European and 3 out of 7 (42.3%) Asian samples, respectively. Serotyping revealed that discordant viral load results between Amplicor 1.5 and Quantiplex 3.0 occurred within samples from all subtypes (A-E). Despite the improvements made to both the Roche Amplicor and the Chiron Quantiplex assays discordant results were detected between the two assays in 25.7% of cases. In a substantial minority of patients there were major discrepancies between the two assays that were not explained by HIV subtype differences.

؉ and CD8 ؉ T lymphocyte responses. In contrast, subjects with virologic failure associated with declining CD4 ؉ T lymphocyte counts had substantially weaker HIV-specific CD4 ؉ T lymphocyte responses and exhibited a trend towards weaker HIV-specific CD8 ؉ T lymphocyte responses. Importantly the CD4 ؉ response was sustained over periods as long as 11 months, confirming the stability of the phenomenon. These correlative data lead to the testable hypothesis that the consequences of viral recrudescence during continuous ART are modulated by the HIV-specific cellular immune response.

The prevention of mother to child transmission of HIV-1 by zidovudine monotherapy is well known, but increasingly combination anti-retroviral therapy is prescribed during pregnancy. In this prospective study, 19 pregnant women with human immunodeficiency virus-1 (HIV-1) infection who elected to take anti-retroviral therapy during the second and third trimesters were treated with zidovudine or zidovudine plus lamivudine. Fourteen women treated with zidovudine monotherapy had a mean 0.3 log(10) reduction in viral load and a mean 52 x 10(6)/L (17%) increase in CD4+ lymphocytes at delivery compared with pre-treatment samples. Genotypic mutations associated with decreased susceptibility to zidovudine were detected in 2 of 10 women at delivery. Five women with more advanced HIV-1 infection were treated with zidovudine plus lamivudine and a mean 1.5 log(10) reduction in viral load together with a mean 30 x 10(6)/L (33%) increase in CD4+ lymphocytes was observed in this group. However, four of five women in the dual therapy arm had the M184V mutation in the reverse transcriptase gene associated with decreased susceptibility to lamivudine at delivery. We conclude that zidovudine plus lamivudine reduced HIV-1 plasma viraemia to low levels in pregnant women with advanced HIV-1 disease but the rapid development of genotypic resistance to lamivudine indicates that additional therapy is required both for the long-term benefit of the mothers and to prevent the development of resistant virus that may be transmitted to the infant.