Background: We investigated the driving forces of the HIV-1 epidemic in Guangxi, China, by examining virologic and epidemiologic data in the IDUs of Binyang and Pingxiang, major urban areas along 2 separate epidemics in the Province.

Methods: A cohort of longitudinally followed drug users sampled in September and October 2000 in the cities of Pingxiang (n = 265) and Binyang (n = 317) were investigated. Sera were tested for HCV by EIA and for HIV by EIA, western blot and were possibly positive for HIV RNA by Roche Amplicor v1.5. Sequence analysis was performed on the gag-pol region (HXB2 nt 1850–3005) and C2-V4 env (HXB2 nt 6704–7626).

Results: The HIV prevalence was 25% and 19% with incidence rates of 5.2 and 8.0 infections per 100 person years in Pingxiang and Binyang, respectively. HIV viral loads showed no statistical difference by subtype or region. Sequence data demonstrated that 93% of the infections in Pingxiang were CRF_01 AE (AE), while 96% in Binyang were CRF_08 BC (BC). One (1) recently infected subject in Pingxiang was infected with BC, and one chronically infected subject had evidence of a new recombinant virus with an AE gag-pol sequence and a BC env. The 2 chronically AE-infected subjects in Binyang were of a different lineage than the epidemic in Pingxiang. Inter-subject distances for gag, protease, RT and C2, V3, C3, V4 of env were 2.05 (±1.78), 1.37 (±1.43), 1.16 (±1.27), 3.23 (±1.26), 4.99 (±2.22), 7.91 (±4.14), 7.78 (±5.10) for AE and 1.02 (±0.81), 0.34 (±0.68), 0.69 (±0.55), 1.61 (±0.99), 0.00 (±0.12), 4.95 (±2.18), 4.77 (±7.25) for BC and statistically different for all regions except V4. Epidemiological data shows different mechanisms for the spread of the 2 epidemics. The BC epidemic in Binyang was clearly one borne by parenteral transmission. On the other hand, the AE epidemic in Pingxiang has 31% of the infected subjects admitting to sex for reasons “other than love”; 2 HIV-infected individuals being HCV^- and 33% of the infected individuals purchasing their drugs outside the city.

Conclusions: The epidemic in Binyang is similar to that seen previously described IVDU-based epidemics with a strong founder effect and little variation in V3. On the other hand, the epidemic in Pingxiang could have multiple introductions of the AE epidemic into the city and greater spread through sexual transmission resulting in greater variation in V3 than typically seen in purely parenterally based epidemics.