Background:  The extent that men who have sex with men (MSM) with transmitted HIV drug resistance (TDR) contributes to ongoing transmission is unknown. We explore possible sources of recently acquired TDR through a phylogenetic and epidemiological analysis of a large MSM open-cohort.

Methods:  Between 2000 and 2006, 1144 HIV-infected MSM from a geographically contained population had infection stage at diagnosis and a dated therapy history documented. Of these, 859 (75%) had HIV pol sequences available (obtained from plasma virus or pro-viral DNA) from which drug resistance mutations were identified. Recent HIV infections were defined as:  anti-HIV positive test within 6 months of a negative test; STARHS; p24 antigen⁺/antibody^– status and limited Western blot. Patients with drug resistance were categorised as:  TDR (recent), TDR, (non-recent) or acquired. Using incidence estimates and epidemiological data, and through phylogenetic and conservative genetic distance measures, the single most likely transmitter for each recently infected TDR patient was identified. The transmitter was categorized as having acquired resistance or TDR (recently or non-recently acquired) at the time the likely transmission took place (dated using the estimated infection date of each recent TDR case). 

Results:  Overall, of 859 MSM, 178 (20.7%) had HIV drug-resistance mutations, of which 13 (7%) had recent TDR. The remainder had non-recent TDR (31 = 17%), acquired resistance (74 = 41%), but for 60 cases (34%) the origin of resistance could not be categorized. Of the 13 with recent TDR, the most likely transmitter was identified in 4 (31%) cases, 2 of which had non-recent TDR and 2 had acquired resistance. Around the estimated time of transmission, those untreated transmitters with TDR had viral loads of >20,000 copies/mL and those with acquired resistance had interrupted therapy with viral loads of >550,000 copies/mL. The most likely transmitter could not be identified for 9 (69%) MSM (with recent TDR, indicating a substantial proportion may have been derived from MSM with undiagnosed infection.

Conclusions:  This exploratory analysis identifies undiagnosed infection, together with high viral load, as risk factors for the onward transmission of resistant strains. Our data support earlier HIV testing and initiation of ARV as public health measures.