Of 2,507 HIV-positive patients seeking medical care at National Taiwan University Hospital (NTUH) between 1994 and 2010, 1,773 (70.7%) were MSM, 480 (19.1%) were heterosexuals, and 187 (7.5%) were IDUs. After the 187 IDUs were excluded, anti-HCV was positive at baseline in 6.5% (102/1,576) of MSM and 8.6% (36/420) of heterosexuals. The present study included a total of 892 patients (35.6%), including 731 MSM and 161 heterosexuals, with negative anti-HCV at baseline and at least one anti-HCV test result during follow-up whose only stated risk of HIV infection was sexual contact. Compared with 1,428 (2,507 − 187 − 892 = 1,428) non-IDU patients who were not included in this study, the 892 study subjects were younger (mean, 41 versus 42 years), less likely to be homosexuals (82.0% versus 73.0%), and more likely to be tested for syphilis (86.3% versus 77.7%) and had a higher plasma HIV RNA load (4.81 versus 4.64 log10 copies/ml) (all P < 0.05); no statistically significant differences were noted in HBsAg seropositivity, RPR titers, and CD4 counts (data not shown).
During the total follow-up duration of 4,270 PY, 30 patients (3.36%) had HCV seroconversion, with an overall incidence rate of 7.03 per 1,000 PY. The rate increased from 0 per 1,000 PY (0/283 PY) in 1994 to 2000 and 2.29 per 1,000 PY (3/1,309 PY) in 2001 to 2005 to 10.13 per 1,000 PY (27/2,667 PY) in 2006 to 2010 (P < 0.05). After exclusion of 161 heterosexuals, the overall incidence rate of HCV seroconversion was 9.25 per 1,000 PY (28/3,025.9 PY) in 731 HIV-infected MSM, which increased from 0 per 1,000 PY in 1994 to 2000 and 3.49 per 1,000 PY (3/858.79 PY) in 2001 to 2005 to 12.32 per 1,000 PY (25/2,029.38 PY) in 2006 to 2010 (P < 0.05).
Compared with 862 patients without HCV seroconversion, 30 HCV seroconverters were more likely to have recent syphilis (43.5% versus 9.4%; P < 0.001), whereas no statistically significant differences existed between the 2 groups in terms of age, gender, HIV transmission routes, HBsAg seropositivity, baseline RPR titers, CD4 counts, and plasma HIV RNA loads (). After adjustment for age and HIV transmission routes, recent syphilis remained an independent factor associated with HCV seroconversion (adjusted odds ratio, 7.731; 95% confidence interval, 3.131 to 19.086; P < 0.001) in multivariate analysis ().
Comparisons of demographic and clinical characteristics of 30 patients with HCV seroconversion and 862 patients without HCV seroconversion
Multivariable analysis for factors associated with HCV seroconversiona
In the nested case-control study, 56 nonseroconverters who had similar observation durations were identified as controls for 30 seroconverters. Seroconverters were less likely to have baseline CD4 < 200 cells/μl but more likely to have recent syphilis acquisition and abnormal aminotransferase levels (all P < 0.05) (). There were no statistically significant differences between the 2 groups in terms of HIV transmission routes, HBsAg seropositivity, VDRL titers, total bilirubin levels, CD4 counts, plasma HIV RNA loads, and receipt of HAART at the last follow-up.
Comparisons of clinical characteristics of 30 HCV seroconverters and 56 nonseroconverters in the nested case-control study
Among the 30 seroconverters, 21 (70%) had HCV viremia, with 7 strains of genotype 1b, 11 of genotype 2a, 2 of genotype 3a, and 1 of genotype 6a (). Phylogenetic analysis was performed to determine their relationship with prevalent HCV sequences amplified from other HIV-infected MSM (n = 33), heterosexuals (n = 8), and IDUs (n = 55). Some of the IDU sequences were excluded from analysis due to the size of the phylogenetic tree; such exclusion did not have any impacts on the outcomes (data not shown). As shown in and in Fig. S1 in the supplemental material, 18 of 21 incidents of HCV infections (85.7%) clustered with at least one other sequence in the tree, resulting in 3 transmission pairs (pairs 1 to 3) and 4 clusters (clusters 1 to 4). The majority of cluster sequences were derived from MSM, except one derived from a heterosexual in pair 3. None of these sequences clustered with sequences amplified from IDUs. Three transmission pairs and cluster 1 were observed within genotype 1b. Clusters 2 and 3 were within genotype 2a, and all patients in those two clusters were recent HCV seroconverters. Cluster 4, consisting of 3 patients, was noted within genotype 3a. Two consecutive sequences were amplified from one of the study subjects, MSM8924, whose first sequence, 2008MSM8924, belonged to genotype 1b, whereas the second sequence, 2008MSM8924-2, amplified from a specimen collected 4 months later, was genotype 2a and belonged to cluster 3.
Fig 1 Phylogenetic analysis of NS5B sequences amplified from HCV seroconverters. The 21 HCV sequences amplified from seroconverters are labeled in yellow. The 96 prevalent HCV sequences amplified from men who have sex with men (n = 33), heterosexuals (n = 8), (more ...)