It has been previously noted that there is discordance

in the spectrum of resistance-associated mutations observed at transmission, compared with those emerging during ART in treated individuals [6,7]. The key lamivudine mutation, M184V, which is the most commonly observed mutation in treated patients, is seldom seen in viruses from untreated patients, including those who have other drug resistance-associated mutations. This has been thought to be attributable selleck inhibitor either to it causing reduced transmissibility of the virus or to the rapid loss of this mutation in the absence of treatment as a result of its impact on viral fitness [6]. Standard resistance testing on plasma is limited to detecting viruses present as majority populations (>20%) within the viral quasispecies. By contrast, some variants may only be present in low proportions, and thus avoid detection.

A number of studies in ART-naïve patients have identified drug-resistant variants found only with Fluorouracil concentration more sensitive assays capable of detecting subpopulations within the virus population with a limit of sensitivity of between 0.001 and 2% [8,9]. The recent studies of Metzner et al. [8] have demonstrated that minority variants of drug-resistant viruses, which were detectable at baseline using only sensitive minority species assays, can outgrow and become the major viral population, leading to virological failure in patients receiving first-line ART. Here we report the prevalence of drug resistance mutations detected with sensitive allele-specific minority assays, compared with genotyping by population sequencing, in an undiagnosed HIV-infected UK population. Our findings suggest a substantial underestimate of drug resistance by currently utilized assay systems. A panel of archived sera collected during 2003 and 2006 from 165 HIV-seropositive homosexual men attending sexual

health clinics in England, Wales and Northern Ireland as part of an ongoing unlinked anonymous serosurvey of HIV infections (GUMAnon) was used in this study Amino acid [10,11]. Eighty-nine samples from 2003 and 76 from 2006 were tested. The GUMAnon survey uses serum collected for routine syphilis tests, and its design allows the exclusion of specimens from subjects with a self-reported HIV infection. The GUMAnon survey has ethical approval for collection and unlinked anonymous testing of specimens. Specimens were selected for testing on the basis of those with sufficient remaining volume for plasma extraction and represented all of the sera available for testing from the collection period in our archive. Participants were all subtype B-infected (as determined by this study) and were designated as having recent (<6 months) or long-standing infections by serological incidence testing [12], using the Vironostika assay (Biomerieux, Basingstoke, UK).