Background:  Non-B HIV subtypes have been associated with heterosexual transmission since the onset of the UK epidemic. By 2000, due to increasing immigration from Sub-Saharan Africa, non-B subtypes represented more than 75% of heterosexual infections and a UK origin was indicated in only 10% of these. We analyzed cases of subtype A and C infection from a high-density sample (25,631 patients) of the entire UK HIV-infected population using recently developed techniques in phylogenetics to infer the pattern and dates of transmission in this group.

Methods:  Partial pol sequences from 11,071 UK patients with non-B subtypes (including 6096 subtype C and 1581 subtype A) were collected during routine clinical treatment. Genetic distance was used to select the subset of patients with a close link (<4.5%) to at least one other individual. Bayesian Markov chain Monte Carlo phylogenetics was used to define clusters, which were analyzed individually using a relaxed molecular clock in BEAST. Transmission dates were estimated as the most recent common ancestor of sequences found in different patients.

Results:  Of the total, 1372 (23%) subtype C-infected and 367 (23%) subtype A-infected patients showed a direct link to at least 1 other sequence. Transmission dates from time-scaled trees showed that 80% of transmissions have occurred since 1995 and 50% since 2000. Minimum transmission intervals were also estimated, for which the median was 27 months. This was much longer than previously described for subtype B sequences among men who have sex with men (MSM) in the United Kingdom (median 14 months). Only 1% of non-B transmissions occurred within 6 months of infection compared with 25% for subtype B. Mean CD4 count at treatment did not differ significantly between subtype A (219 cells/µL) and B (228 cells/µL) and was not greatly lower for subtype C at 205 cells/µL, so the difference was not due to earlier treatment among MSM. However, only 85 C, (6%) and 53 A strains (14%) were in UK-based clusters with more than 10 individuals (vs 25% for subtype B).

Conclusions:  Phylodynamic analysis identified a large number of non-B subtype transmissions involving UK sequences since 2000, suggesting many to have occurred within the United Kingdom. However the numbers in large clusters is lower and the inter-transmission interval for non-B subtype strains (predominantly found in heterosexuals) is much longer than for subtype B in MSM suggesting the epidemics have very different transmission dynamics.