Active or Prior GB Virus C Infection Does Not Protect against Vertical Transmission of HIV in Coinfected Women from Tanzania

GB virus C (GBV-C) is a flavivirus that is closely related to hepatitis C virus; however, it does not replicate in hepatocytes and has not been associated with any specific disease (reviewed in [1]). GBV-C viremia can persist for decades without apparent illness, or the virus may be cleared with the formation of antibodies to envelope glycoprotein E2 [2, 3]. Recent studies found that coinfection with GBV-C in HIV-infected patients was associated with an increase in the duration of survival, independent of HIV treatment received, baseline CD4⁺ T lymphocyte count, age, sex, race, and mode of transmission (reviewed in [4]). GBV-C viremia was also shown to slow the progression to AIDS in patients with coinfection. GBV-C replicates in CD4⁺ T lymphocytes in vitro [5], and coinfection of PBMCs with GBV-C and HIV has resulted in decreased HIV replication [6]. Although the mechanism of interaction between GBV-C and HIV is not completely understood, there is some evidence that GBV-C replication alters chemokine expression, which may contribute to the inhibition of HIV replication [7].

The prevalence of GBV-C viremia among blood donors is 1%–2% in the United States and up to 4% in some developed countries [1]. Among HIV-infected individuals, the prevalence of GBV-C viremia may be as high as 40% [8]. Parenteral transmission of GBV-C is well recognized, and recent studies suggest that sexual transmission occurs frequently as well [9]. Several studies have found a high rate of vertical transmission of GBV-C (60%–80%) from viremic mothers [10, 11], and small studies have evaluated whether coinfection with HIV affects the vertical transmission of GBV-C. In one of these studies [10], 34 mothers with detectable GBV-C RNA levels were evaluated. Eight of these women were HIV positive. There was not a statistically significant difference in the rate of vertical transmission of GBV-C between babies who were born to mothers with HIV coinfection (50%) and those who were born to mothers without HIV infection (65%). Another study evaluated vertical transmission of GBV-C in HIV/GBV-C–coinfected women and found a rate of vertical transmission of GBV-C of 45%, which is similar to the rate seen in women without HIV infection [11]. The effect of maternal GBV-C viremia on the vertical transmission of HIV, however, remains unknown. Thus, we evaluated the effects of GBV-C infection on the vertical transmission of HIV using archived plasma samples obtained from HIV-positive women in Dar-es-Salaam, Tanzania, from 1992 through 1993.

Methods. This was a retrospective cohort study using archived serum samples obtained for a study on the vertical transmission of HIV during the period of 1992–1993. The institutional review boards at Duke University (Durham, NC) and the University of Iowa (Iowa City, IA) approved the study. Women attending an antenatal clinic in Dar-es-Salaam, Tanzania, were addressed by a Swahili-speaking nurse and invited to participate in screening for the HIV vertical transmission study. Consenting women were interviewed privately to reconfirm their consent to having an HIV test performed, and they subsequently were tested for HIV by ELISA. Blood samples were stored at -20°C. Women who tested positive for HIV infection twice by ELISA were enrolled if they consented to continue with the study. Blood samples were obtained from their infants at birth and 6 weeks, 3 months, 6 months, 9 months, 12 months, 15 months, and 18 months after birth for determination of HIV RNA levels by RT-PCR. None of the study subjects received antiretroviral therapy.

All serum samples were identified by a unique code, and laboratory personnel were blinded to clinical data. RNA was extracted from 200 µL of serum using a previously described guanidinium-isothiocyanate extraction method [6]. One-fourth of this RNA preparation was used in nested RT-PCR reactions to amplify GBV-C RNA using primers from the E2 coding region of the genome. Specifically, primers were as follows: outer sense, 5′-GDC GYG AYT CGA ARA TMG AYG-3′ (nt 1484–1505); outer antisense, 5′-AAG ATC AAC GGG ACC AGC CGT GCC TCA-3′ (nt 2229–2251); inner sense, 5′-GAT ATC GAA RAT MGA YGT GTG GAG-3′ (nt 1490–1512); inner antisense, 5′-TTA GGT ACC GCC TCA GCC AGC TTC AT-3′ (nt 2219–2235) (sequence numbering based on the full-length Iowa GBV-C isolate; GenBank accession no. AF121950 [5]). Reaction conditions and the protocol for PCR product identification were described elsewhere [6]. Negative and positive controls were included with each sample undergoing PCR testing. The presence of antibodies to GBV-C envelope glycoprotein E2 was determined by ELISA [12], using the µPlate anti-HGenv test (Roche Diagnostics).

Statistical analyses were completed using Stata software, version 6.0/7.0 (Stata). Binomial 95% CIs were calculated using exact methods, when appropriate. P values are 2-tailed and were estimated using the Wilcoxon rank sum test.

Results. Two thousand three hundred twenty-nine women were screened, and 320 (13.8%) were found to have antibodies to HIV. Of these women, 215 HIV-positive women enrolled in the study and had 131 live births. Of the 215 HIV-infected women enrolled in the vertical transmission study in Tanzania, 186 had serum samples available for GBV-C testing. All but 1 of the infants in this study were delivered vaginally. Ninety-eight (53%) of the 186 women had evidence of past or active GBV-C infection. Forty-four women (24%; 95% CI, 18%–30%) had detectable GBV-C RNA, and 54 (29%; 95% CI, 24%–37%) had anti-E2 antibodies. Two patients had both GBV-RNA and anti-E2 antibodies.

The rate of vertical transmission of HIV according to GBV-C status is shown in figure 1. HIV-positive women with GBV-C viremia transmitted HIV to 25 (28%; 95% CI, 12%–49%) of their infants, compared with 26 (23%; 95% CI, 9%–44%) of the infants from women with anti-E2 antibody and 77 (26%; 95% CI, 17%–37%) of the infants from women without active or past GBV-C infection. There was no statistically significant difference in the rates of vertical transmission of HIV.

Figure 1

Rate of vertical transmission of HIV, according to the hepatitis GB virus C (GBV-C) status of the mother. P = .60, for women with GBV-C viremia versus those without GBV-C infection. P > .99, for women with GBV-C antibodies versus those without GBV-C infection. +, Positive; -, negative.

Discussion. GBV-C viremia has been shown to be associated with an increase in the duration of survival among HIV-infected individuals, independent of HIV treatment, baseline CD4⁺ T lymphocyte count, age, sex, race, and mode of transmission (reviewed in [4]). Although the interactions between GBV-C and HIV are incompletely understood, there is evidence that GBV-C replication inhibits HIV replication [6, 7]. We studied the prevalence of GBV-C infection and the effects of coinfection with GBV-C on the vertical transmission of HIV in Tanzanian women who did not have access to antiretroviral therapy.

The prevalence of GBV-C viremia in HIV-infected, pregnant Tanzanian women was 24%, and an additional 30% had detectable antibodies to envelope protein E2, indicating past GBV-C infection and clearance. Similar to previous studies [8], the prevalence of GBV-C viremia in HIV-infected women was significantly greater than the prevalence of GBV-C reported in the general population. However, the prevalence of GBV-C viremia in this study was lower than that seen in some HIV-infected cohorts. For example, a study of 362 HIV-infected individuals from the midwestern United States found a prevalence of GBV-C viremia of 40%, regardless of risk factors for HIV infection [6], and a retrospective evaluation of samples obtained from 271 HIV-infected individuals in the Multicenter AIDS Cohort Study (MACS) found a prevalence of GBV-C viremia of 39% [12].

We assume that the vast majority of HIV infections in Tanzania were transmitted sexually. GBV-C has been shown to be transmitted sexually as well as parenterally [10, 11]; therefore, this probably does not account for the difference in prevalence. All of the samples in this study were obtained in the same manner and frozen until use, so there should not have been technical difficulty in detecting GBV-C RNA in the samples. The lower overall rate of viremia and E2 antibody positivity may reflect an increased rate of clearance of GBV-C RNA without development of E2 antibody in HIV-positive individuals. In the MACS cohort, for example, 12 of 61 participants who were GBV-C RNA positive within 18 months of HIV seroconversion cleared their GBV-C viremia during 4–5 years of follow-up. Seventy-five percent of these patients did not develop E2 antibody, and these patients were 5.87 times more likely to die than were those who were persistently viremic [12]. Another factor that may contribute to the difference in the prevalence of GBV-C viremia seen in this cohort is that, presumably, the Tanzanian women had non–clade B HIV and either GBV-C genotype 1 or 5. It is unknown how either HIV or GBV-C diversity might affect the detection of GBV-C infection or the interaction between GBV-C and HIV. Thus, this study may not be applicable to all HIV-positive pregnant women.

Previous studies found a high rate (45%–80%) of vertical transmission of GBV-C, but none of these studies evaluated the effects of coinfection with GBV-C on the vertical transmission of HIV. In this cohort, neither maternal GBV-C viremia nor E2 antibody was associated with a high level of protection against vertical HIV transmission. However, because of the relatively small number of subjects with an adequate volume of serum available for evaluation of GBV-C RNA or anti-E2 antibodies, this study did not have sufficient power to detect small differences in HIV transmission rates. With 51 coinfected women and 77 women with HIV infection alone, as well as a baseline rate of vertical transmission rate of 26%, then (with an α of 5% and using Fisher's exact methods, given the small sample size) the study had 80% power to detect an absolute difference of 20%. Nevertheless, these data do not support a protective effect of GBV-C infection on vertical transmission of HIV in Tanzanian women. The possible effect of vertical transmission of both HIV and GBV-C on long-term infant survival is currently under study.

• Received September 2, 2003.
• Accepted November 16, 2003.

• © 2004 by the Infectious Diseases Society of America