Baseline characteristics of ATAHC participants, stratified by mode of acquisition, are shown in . Overall, 167 subjects were enrolled, with 4 failing to return after screening, resulting in a total study population of 163 subjects.
Participant Characteristics, Stratified by Mode of Acquisition
One hundred nineteen subjects (73%) had IDU identified as the most likely route of infection, 29 (18%) were most likely infected through sexual exposure, and in 15 cases (9%), ”other” or “unknown” routes of infection were identified (). Of the 29 sexually acquired cases, 4 involved HIV–uninfected women, all of whom had a known HCV–infected partner and no history of IDU. Of the 25 men exposed sexually, only 2 were HIV uninfected, whereas 23 (92%) were HIV infected. Of the 2 HIV-uninfected men, 1 was infected through same-sex (MSM) exposure with partner(s) of unknown status, and 1 was exposed through sexual contact with a woman known to be HCV infected. All 23 HIV-infected subjects were sexually exposed through MSM exposure. Of the 15 cases with other routes of exposure identified, 2 were suspected infections through medical procedures, 1 was associated with occupational exposure, 2 were associated with nonoccupational needlesticks, 1 was associated with assault, and 4 cases had unknown routes of exposure. In 3 cases, either MSM sexual exposure (n = 2) or IDU (n = 1) were disclosed by the subject on answering the risk assessment survey but were not identified by the clinician on the primary case report form.
Clinical, demographic, and behavioral characteristics, stratified by mode of acquisition, are shown in . Individuals exposed through sexual contact were more often male and HIV infected. Aside from expected differences in drug use characteristics, other notable differences between individuals with sexual and IDU-related acquisition were observed, including differences in education level, employment, prison experience, social functioning, and depression. Clinical characteristics, however, were generally similar between the groups (.)
Participants were stratified further on the basis of both HIV status and mode of acquisition (IDU vs sexual exposure) (). Because the group of HIV-uninfected subjects with sexual acquisition comprised only 6 cases, this group was excluded from analysis, leaving 3 groups; the HIV-uninfected IDU group (n = 96), the HIV-infected IDU group (n = 23), and the HIV-infected sexual acquisition group (n = 23). Characteristics of these 3 groups are given in . Marked differences were seen between the HIV-uninfected IDU group and the HIV-infected sexual acquisition group, similar to those described in . Indices of social stability, such as higher education, home ownership, and employment, were all highest for the sexual acquisition group, followed by the HIV-infected IDU group and then the HIV–uninfected IDU group. The social functioning score, a global indicator of social stability in which higher values reflect greater instability, demonstrated that 50% of HIV-uninfected individuals had scores >14, compared with 22% of HIV-infected IDU cases and 13% of HIV-infected sexual acquisition cases.
Participant Characteristics Stratified by HIV Status and Mode of Acquisition
A phylogenetic tree was constructed using 116 E1/HVR1 sequences from 112 screening samples and 4 cases of reinfection (). Of the 163 samples from enrolled subjects, sequences could not be obtained for 11 screen samples (6.7%) because of technical difficulties and for an additional 40 screen samples (24.5%) because of the low level or absence of viremia (virus load, <615 IU/mL). The majority of genotypes were 1a (53 [47%]) and 3a (50 [44%]), with smaller numbers of 1b (3), 2a/b (6), and 4a (1). Of the 112 sequences subjected to phylogenetic analyses, 77 were obtained from HIV–uninfected subjects and 35 were from HIV-infected subjects, representing 68% and 70% of each group, respectively. Phylogenetic analysis using a bootstrap value of >98% identified 4 clusters and 3 homologous pairs of virus, containing a total of 23 viruses (20% of the total sequenced) (). Of these 23 viruses, 18 were from HIV-infected subjects (51% of HIV-infected subjects sequenced) and 5 were from HIV-uninfected subjects (8% of HIV–uninfected subjects sequenced; P < .001). The vast majority (19 of 23) of these viruses were in genotype 1a clusters and involved both IDU-acquired (n = 15) and sexually acquired (n = 7) cases. Details of subjects falling into the clusters and pairs by HIV status and mode of acquisition are given in . HIV-infected clusters contained individuals with both sexual and IDU-related acquisition (clusters 4–6). One separate pair of HIV–uninfected IDUs was identified (cluster 2). Clusters 1 and 3 contained both HIV-infected and HIV-uninfectedindividuals. Cluster 1 contained 1 HIV–uninfected patient who acquired HCV infection through IDU, one HIV–uninfected subject with “other” identified as the most likely route of transmission, and 1 HIV-infected subject with HCV reinfection acquired through sexual exposure. In cluster 3, 2 HIV-infected IDUs clustered with an HIV–uninfected MSM infected through sexual exposure. Although these clusters 1 and 3 contained participants with differing HIV status, further exploration revealed that both HIV-infected subjects in cluster 3 identified their sexual orientation as MSM, as did 2 subjects in cluster 1—namely, the HIV–uninfected subject with IDU-acquired HCV infection and the HIV-uninfected subject who acquired HCV infection through an “other” route. Thus, irrespective of the most likely mode of HCV acquisition and HIV status, all subjects included in clusters or pairs in this analysis were MSM (with the exception of the 1 pair of female HIV-uninfected IDUs).
Figure 1. A, Phylogenetic analysis of 116 E1/HVR-1 sequences (657 bp) from 112 hepatitis C virus (HCV) RNA–positive patients demonstrating clustering of subjects enrolled in the Australian Trial in Acute Hepatitis C. Clusters are numbered 1–7. Spheres (more ...)
Clinical and Demographic Characteristics of 23 Clustering Subjects
Molecular Clock Analysis
The degree of geographic dispersal and the plausible origin of HCV isolates in clusters or pairs were estimated on a real time scale using Markov Chain Monte Carlo approach to calculate the year of the most common recent ancestor (). The hierarchy presented in the phylogeographic tree indicated that the earliest divergence events occurred in New South Wales around 1989, and all recent migration events for monophylogenetic clusters or pairs occurred after 1998, after the introduction of HAART. Most sequences originated from New South Wales and Victoria with the exception of early migration (1991) of a genotype 3a virus from New South Wales to South Australia (SA). The number of migration events prior to the cluster formation ranged from 2 to 7 (mean ± standard deviation, = 4.7 ± 1.9).
In addition, we examined whether there was evidence for different timings of transmission networks in HIV-uninfected and HIV-infected ATAHC subjects using markers of evolutionary tree diversity. Bushiness models were constructed with subject sequences using BEAST software. AvgDL was calculated for both HIV-infected subjects and HIV-uninfected subjects (). The AvgDL in HIV-infected subjects was significantly lower than that in HIV-uninfected subjects for both genotype 1 (13.00 vs 19.98; P < .001) and genotype 3 (11.62 vs 20.42; P < .001) evolutionary trees.
Analysis of Phylogenetic Tree Evolutionary Characteristics, by Human Immunodeficiency Virus (HIV) Infection Status