|Home | About | Journals | Submit | Contact Us | Français|
GBV-C has been associated with a better prognosis of HIV-1 disease in adults. Little is known about prevalence and interaction between GBV-C, HIV-1, and/or HCV in hemophiliac children and adolescents.
A well characterized cohort of HIV-1-infected and -uninfected hemophiliac children and adolescents followed in the Hemophilia Growth and Development Study (HGDS) were evaluated using qRT-PCR to detect GBV-C RNA in samples from baseline and last follow-up visit.
HIV-1-infected (n=202) and -uninfected (n=119) patients had a low prevalence of GBV-C infection at baseline (0.9 and 0%), which increased at time of last follow-up visit to 25.2% and 26.3%, respectively. In addition, at the time of the follow-up GBV-C measurement, those GBV-C-infected had been followed longer and had higher CD4+ cell counts and lower HIV-1 viral loads than those GBV-C-uninfected. These beneficial effects of GBV-C were no longer significant after controlling for CD4+ cell count and HIV-1 RNA at baseline. HCV RNA clearance was more common amongst those who were not GBV-C infected than those who became GBV-C-viremic.
This study confirms a positive association of GBV-C with milder course of HIV-1 infection. GBV-C infection was associated with a higher likelihood of persistent HCV infection.
Almost all haemophiliac patients were infected with the hepatitis C virus (HCV) when receiving clotting factor concentrates before the initiation of heat treatment in 1985 [1,2], and many were also infected with the human immunodeficiency virus type 1 (HIV-1). There has also been increasing interest in another member of the flaviviridae family, the GB virus C (GBV-C), which is the closest relative to HCV in man. While HCV has been associated with liver disease and impaired outcome in HIV-1-infected patients [3,4], GBV-C has not been connected to any disease [5,6] but has been shown to be associated with an attenuated course of HIV-1 disease [5,7,8].
GBV-C viremia exists in up to 40 percent of HCV and/or HIV-1-infected adults [9–17]. However, until now little is known about the prevalence of GBV-C in hemophiliac children or its influence on HCV-infection. This study investigates the prevalence and interaction between GBV-C, HIV-1 and HCV in a cohort of HCV/HIV-1 co-infected and HCV mono-infected hemophilic children and adolescents.
The Hemophilia Growth and Development Study (HGDS) was a United States based multicenter study that enrolled patients from 1989 until 1990. Patients' age at entry ranged from 6 to 19 years with samples collected at baseline and 6 month intervals through 1997. Assessments were made of medical histories, physical examinations were performed, and CD4+ T cell counts were measured in both the HIV-1-infected and -uninfected patients. Blood samples were processed within 24 hours for cryopreservation of cells, plasma, and serum. During the course of follow-up, antiretroviral therapy was prescribed at the discretion of the primary providers; however, potent triple drug therapy only became widely available during the last year of follow-up with only 9 patients ever having received such treatment. During the course of follow-up 57 progressed to AIDS and 69 died. AIDS progression was defined on the CDC Surveillance Case definition for AIDS . All parents or legal guardians as well as all patients gave informed consent prior to any study related intervention. The HGDS has been approved by the human subjects committees of the collaborating institutions.
The plasma samples chosen for GBV-C analysis included the earliest one available, which was defined as baseline for this substudy, independent of the actual time in relation to enrollment into the HGDS study and the last sample available during follow-up. For those HIV-1-infected, samples were available at baseline for 202 of 207 enrolled with follow-up samples for 191. Amongst those HIV-1-uninfected, samples were tested at baseline in 119 of 126 total enrolled with follow-up samples tested for 114. Stored plasma samples had previously been tested for HIV-1 and HCV RNA levels by bDNA assay as well as HCV genotypes as described earlier .
In the present study samples of HIV-1-infected and -uninfected patients were tested for GBV-C RNA. Viral RNA was isolated from plasma using Viral RNA Mini Kit (Qiagen, Hilden) following the instructors manual. GBV-C RNA was quantified by one-step real time PCR on a LightCycler 1.0 (LightCycler RNA Amplification Kit; Roche Diagnostics, Mannheim). PCR was performed using Primers (forward: 5'- AATCCCGGTCAYCYTGGTAGCCACT-3', reverse: 5'- CCCCACTGGTCYTTGYCAACTC-3') and a taqman probe (5'-6FAM-AATAAGGGCCCGACGTCAGGCTCXT-PH-3') compatible to the 5'-untranslated region (UTR) of GBV-C. PCR protocol started with a 30 minutes reverse transcription cycle at 55°C, followed by PCR: 1 minute denaturation for initialization. 60 PCR cycles each with a 5 seconds denaturation at 94°C, 15 seconds primer annealing at 56°C and 15 seconds amplification at 72°C were carried out. The sensitivity of the PCR was 500 copies/mL.
Summary statistics were computed as frequency and percent for categorical outcomes, and mean and standard deviation (SD) for continuous outcomes. For the follow-up analysis, the tests of differences between the GBV-C-infected (GBV-C RNA positive) and -uninfected (GBV-C RNA negative) patients at follow-up were performed using a t-test for the difference in means (for age, CD4+ count, log10 HIV-1 RNA, and log10 HCV RNA) and Fisher's Exact test for categorical outcomes, HCV clearance (HCV antibody positive/HCV RNA negative) and use of antiretroviral treatment. The tests for the effect of baseline values on becoming GBV-C viremic were also performed using t-tests for the difference in means (baseline CD4+ count, log10 HIV-1 RNA, and log10 HCV RNA). Longitudinal models were created to assess the effect of being GBV-C-infected at follow-up on CD4+ count (used square root CD4+ count for the analysis), log10 HIV-1 RNA and log10 HCV RNA adjusting for study visit, baseline age, use of antiretroviral treatment, as well as baseline CD4+, HIV-1 RNA and HCV RNA. Time to progression to AIDS or death in the presence of AIDS was analyzed using Kaplan Meier plots and Cox Proportional Hazards models.
At time of baseline, defined as the point of time the first GBV-C testing was done, mean (SD) age for HIV-1-infected patients was 13.9 (±3.3) years and for HIV-1-uninfected patients 11.2 (±3.2) years (p<0.001). In HIV-1-infected patients CD4+ cell counts averaged 382 (±320.0) cells/μL, and median HIV-1 RNA level was 3.4 (±0.8) log10 copies/mL. HIV-1-uninfected patients showed a mean (SD) CD4+ cell count of 957 (±376.9) cells/μL. Mean (SD) HCV RNA levels were significantly higher in HIV-1-infected patients, and 96.5% were HCV RNA positive with a viral load of 6.3 (±1.2) log10 copies/mL versus HIV-1-uninfected patients (5.4±1.2 log10 copies/mL, p<0.0001). Clearance of HCV infection was defined as HCV antibody positive/HCV RNA negative and was observed in 7 of 202 (3.5%) HIV-1-infected versus 19 of 119 (16%) HIV-1-uninfected patients (p=0.0002). Patient characteristics at time of first GBV-C testing for HIV-1-infected and -uninfected patients are summarized in Table 1.
Samples were available at different points of time during the HGDS and were tested for GBV-C RNA. Interestingly, no sample tested positive during the first two years. Two HIV-1-infected patients were positive for GBV-C-RNA at the first point of time tested; however, these samples were at 2.5 and 5 years after initial recruitment with both having the follow-up sample being GBV-C RNA negative.
At follow-up, prevalence of GBV-C viremia was 24.6% in HIV-1-infected and 26.3% in -uninfected patients, however the difference was not significant. There was also no difference in GBV-C viral load between those HIV-1-infected and -uninfected (5.5±0.7 versus 5.6±1.1 log10 copies/mL, respectively, p=0.64). GBV-C prevalence among the HIV positive patients differed according to CD4+ cell strata: CD4+ <50 − 14%, CD4+ 50–200 − 28%, CD4+ 200–400 − 18%, CD4+ >400 − 44% (p=0.004).
We observed a statistically significant correlation between years after follow-up and GBV-C prevalence with an odds ratio (95% confidence interval), 1.35 (1.16–1.56), p<0.0001 (Supplemental Figure 1). It was seen that the risk to get GBV-C infected was increased with a longer follow-up.
As only two patients were GBV-C positive at the earliest available sample, we based all the following analyses on the GBV-C results at the time of follow-up testing.
Of the 191 HIV-1-infected patients that had a follow-up sample available for testing, the median time between first and last sample was 4.6 years with a range of 0.4–7.7 years. GBV-C RNA was measured and individuals were grouped into GBV-C RNA positive and negative. Characteristics for these two groups are summarized in Table 2.
At time of the follow-up sample, patients who were found to be GBV-C viremic were on average older than those uninfected (19.2±4.0 versus 17.7±3.1 years, p=0.028). In addition, GBV-C-infected compared to -uninfected patients had significantly higher CD4+ cell counts, mean (SD) of 280 (±275.3) versus 191 (±225.5) cells/μL, p=0.043, and significantly lower mean (SD) plasma HIV-1 RNA levels of 3.4 (±0.8) versus 3.8 (±0.9) log10, copies/mL (p=0.009). In contrast, there was no significant difference in mean (SD) HCV RNA levels between the two groups, 6.2 (±1.5) versus 6.0 (±1.6) log10 copies/mL (p=0.628).
Because of the described effect of GBV-C co-infection being associated with delayed HIV-1 course [5,7,8], progression to AIDS was calculated in relation to time and GBV-C status in the HGDS. Twenty-one patients developed AIDS prior to their first GBV-C measurement, 43 progressed to AIDS between the first and second GBV-C measurement and 127 did not progress to AIDS during the study. Patients who progressed to AIDS prior to the second GBV-C measurement (n=64) as well as patients who had the end of their follow-up (n=99) at the second point of time were excluded from further analysis. After this, 28 patients remained. Seven of 24 (29.2%) GBV-C-uninfected patients and 1 of 4 (25.0%) GBV-C-infected patients went on to develop AIDS. Due to the small number of individuals, no statistical tests were performed to compare these proportions. Additionally, time from second GBV-C measurement to AIDS-related death for all HIV-1 positive patients was examined. The Kaplan-Meier analysis in Figure 1 shows a lower rate of progression to AIDS or death in the presence of GBV-C in those that were found to be GBV-C/HIV-1 dually infected. A Cox Proportional Hazards model resulted in a hazard ratio (HR) of 0.354 and a p-value of 0.045, consistent with the fact that those patients found to be GBV-C infected had higher CD4+ cells and lower plasma HIV-1 RNA levels.
Next we investigated whether baseline characteristics were related to GBV-C status at follow-up (Table 3). At baseline, patients who were later found to be GBV-C-infected had higher mean (SD) CD4+ cell counts, 465 (±284.1) versus 377 (±327.8) cells/μL (p=0.031), and lower mean (SD) HIV-1 RNA levels, 3.2 (±0.8) compared to 3.5 (±0.8) log10 copies/mL (p=0.030) than those never found to be GBV-C-infected. There was no significant difference in HCV RNA levels between the two groups (p=0.209). When analyses were repeated, adjusting for baseline values there was no longer a significant correlation between HIV-1 RNA levels, CD4+ cell counts, and GBV-C infection status. Additionally, survival analysis controlling for baseline HIV-1 RNA level, CD4+ cell counts, HCV RNA level and antiretroviral treatment showed no relationship between GBV-C status and AIDS-related death (data not shown). Furthermore, testing for likelihood of AIDS in relation to GBV-C status at the second GBV-C measurement was done. Seven of the 43 (16.3%) HIV-1-infected patients who progressed to AIDS between baseline and follow-up were GBV-C viremic while 37 of the 127 patients (29.1%) who did not develop AIDS were GBV-C viremic (Chi-square p=0.096).
Because of the correlation between HCV genotype and outcome of HIV-1 infection in a recent study , HCV genotype was examined in relation to progression to AIDS, HIV-1 RNA levels, CD4+ cell counts, and GBV-C status. No statistically significant risk for AIDS progression according to HCV genotype in correlation to GBV-C status was observed (HCV genotype 1 versus others, HR 1.25, p=0.84), nor for AIDS-related death, HR 1.29, p=0.62.
Six of 144 (4.2%) GBV-C negative HIV-1-infected patients were able to clear their HCV infection during the follow-up. In contrast, none of the 47 GBV-C/HIV-1 coinfected patients did. This difference was not statistically significant (p=0.339).
In addition to the influence of GBV-C on HIV-1 infection, a possible effect of GBV-C infection on HCV RNA levels was investigated. Similar to the HIV-1-infected cohort, HIV-1-uninfected patients were grouped according to their GBV-C status at follow-up. The median time between first and second GBV-C measurement was 6.1 years with a range of 1.5–7.3 years. As summarized in Table 4, there were no differences between GBV-C-infected and -uninfected patients in mean (SD) age (17.4±3.5 versus 17.1±3.2 years, p=0.664) or mean (SD) HCV RNA level (6.0±1.3 versus 5.6±1.3 log10 copies/mL, p=0.163), respectively. In contrast, those ultimately found to be GBV-C-infected had higher baseline mean (SD) HCV RNA levels than those not found to be GBV-C-infected (6.0±1.1 versus 5.1±1.2 log10 copies/mL, p=0.001). Comparable to HIV-1-infected patients, those who were never found to be GBV-C-infected were more likely to have cleared their HCV infection: 18 of 84 (21.4%) versus 1 of 30 (3.3%), p=0.023.
Several studies have shown GBV-C viremia to be associated with slower HIV-1 disease progression [5,7,20]. This includes an inverse relationship between GBV-C and HIV-1 RNA levels as well as an increase in GBV-C RNA levels in those starting potent antiretroviral therapy . These observations argue for a causal virus-virus interaction which is supported by in vitro evidence for an attenuating effect of GBV-C on HIV replication [21–24]. Despite that, some authors still argue that the observation only indicates an epiphenomenon . In evaluating the role of GBV-C viremia in this well characterized cohort of HIV-1-infected and -uninfected hemophiliac children and adolescents we identified three findings : 1) low prevalence of GBV-C in both groups during the early phase of the study, increasing to levels more consistent with what has been described in other cohorts at later points of time; 2) milder clinical course of HIV-1 disease in subjects being GBV-C positive at later point of time, which, however, was lost when controlled for baseline factors; and 3) a higher rate of HCV clearance in patients never found to be GBV-C viremic.
The low prevalence of GBV-C viremia at baseline in both groups was surprising. However, after a median of 4.6 years of follow-up for HIV-1-infected and 6.1 years for HIV-1-uninfected patients, prevalence of GBV-C infection was 24.6% and 26.3%, respectively. These later rates are similar to what has been seen in other cohorts of HIV-1- and HCV-infected individuals (15–45% GBV-C-infected) [9–17]. Degradation of GBV-C RNA is an unlikely explanation for the low prevalence at baseline as we were able to detect HCV RNA in GBV-C negative samples and repeated thaw-and-freeze cycles of GBV-C positive plasma did not show loss of GBV-C RNA (unpublished data). Furthermore, we reliably detected HCV RNA in samples from the early cycles (data not shown). HIV-1 and HCV infection in the HGDS cohort is associated with receipt of virus contaminated clotting factors [1,2,26,27]. Low prevalence of GBV-C infection directly after exposure to such clotting factors but increasing during follow-up could suggest that GBV-C is less efficiently transmitted by pooled blood products than HCV or HIV, but is transmitted by other normal contacts or during routine health care. Several studies have shown that HCV transmission can occur in association with increasing numbers of hospital visits [28–31]. If also true for GBV-C it could explain the higher GBV-C prevalence at the end of the study e.g. caused by increased frequency of exposure to clotting factors. The identification of a highly significant correlation between time period of follow-up and increased risk of GBV-C infection would be consistent with this hypothesis and what has been observed in other studies [32–35]. Another possible explanation for our findings is that target cells might become more susceptible to GBV-C as patients age. Some studies evaluated the GBV-C prevalence in children and also found evidence for increase of GBV-C prevalence in older compared to younger children [32,36]. The fact that children do not have fully developed their immune systems may promote HIV-1 replication and depletion of CD4+ cells, whereas a lack of totally differentiated immune cells may impair GBV-C replication and its potential effect on HIV-1 infection. In fact, having higher CD4+ cell counts and lower HIV-1 RNA levels at baseline was predictive of becoming GBV-C RNA positive. Finally, it is possible that hemophiliacs are at somewhat lower risk since pooled sera/plasma can contain anti-E2 antibodies which could neutralize co-existing GBV-C viremia. This would be consistent with other observations that the rate of GBV-C viremia is relatively low in hemophiliacs exposed to GBV-C containing blood products .
We demonstrated that among HIV-1-infected children and adolescents, those that became GBV-C infected at follow-up had higher CD4+ cell counts and lower plasma HIV-1 RNA levels at the time of the follow-up visit than those without detectable GBV-C. However, these associations were no longer significant when controlling for baseline CD4+ cell count or plasma HIV-1 RNA. The lack of improved prognosis after controlling for baseline factors is in line with another study in perinatally HIV-1-infected children . However, it cannot be ruled out that subdetectable GBV-C infection might have already been present and responsible for the better baseline values. Furthermore, it can be discussed if adjusting for CD4+ cell count or plasma HIV-1 RNA CD4 is appropriate as both variables might well be influenced by an underlying GBV-C infection.
HCV infection is very common in HIV-1-infected patients (33–47%) and the clinical course of HCV infection is influenced by HIV-1 [38,39]. In this study we found a difference in the likelihood of clearing HCV depending on HIV-1-infection status (at baseline and follow-up): 31.9% (38/119) of HIV-1-uninfected patients versus 6.4% (13/202) of HIV-1-infected patients. Furthermore, as previously reported from this cohort , HCV RNA levels were lower in those HIV-1-uninfected than -infected patients. These findings can be explained by impairment of cellular immune response due to HIV-1 infection [40,41]. In contrast, there was no relationship between GBV-C infection and HCV RNA level in either the HIV-1-infected or -uninfected patients. However, HCV RNA clearance was significantly (p=0.023) more common amongst those HIV-1-uninfected who were never found to be GBV-C-infected than those that were GBV-C viremic at follow-up. In fact, among the HIV-1-infected patients, HCV clearance was only seen amongst those that were never found to be GBV-C-infected. However, only so few patients clear HCV in the HIV population that this difference was not significant. However, similar trends have been seen in other studies: a French cohort of hemophiliac patients , and in a study in a peritoneal dialysis population . This also fits with some reports of higher sustained viral response for HCV in patients clearing GBV-C viremia during interferon based therapies [9,44]. The relation between HCV persistence and GBV-C viremia may suggest similar effectors of immune response leading to simultaneous control of both viruses. Single HCV proteins have to be shown to block the activation of the immune system and thereby inhibit the efficient elimination of the virus [45–47], and based upon the close phylogenetic relationship between GBV-C and HCV, it is conceivable that such a characteristic could be shared with GBV-C proteins. Recently, IL28B, emerged as predictor of spontaneous and treatment-induced HCV clearance . In one cohort where we could confirm the role of IL28B for HCV clearance, we did not find a role for IL28B in relation to GBV-C clearance .
There are several limitations of this study, including the retrospective nature of the study and the relatively small number of patients. In addition, the small number of patients with AIDS related death precluded statistical tests. The study is further limited by the variability in the sample availability at different points of time. Nevertheless, our study provides new insights into GBV-C, HIV-1 and/or HCV co-infection in hemophiliac children and adolescents.
In summary, our data suggest a better prognosis of HIV-1 disease in patients who are eventually GBV-C infected. It cannot be excluded that GBV-C is a mere marker for a more benign course of disease. Interestingly, absence of GBV-C viremia was associated with a higher rate of HCV clearance.
Correlation between years of follow-up (cycle) and GBV-C status of all patients that were ultimately found to be GBV-C viremic in the HGDS. Black marks indicate the percentage of GBV-C positive samples, while the grey triangles indicate the upper and lower confidence interval (95%). Longer follow-up increases chance of GBV-C infection (p<0.001).
We are indebted to the children, adolescents, and parents who volunteered to participate in the HGDS, and to the members of the Hemophilia Treatment Centers.
Source of Funding: The HGDS was supported by the National Institute of Child Health & Human Development at the National Institutes of Health [1 R01 HD41224]. GBV-C testing was funded by the German “Network of Competence for Hepatitis” (now Deutsche Leberstiftung).
Presentation of data: 8. Research Festival der Universität Leipzig, December 2009, Leipzig, Germany.
Conflicts of Interest None of the authors has any conflict of interest in relation to this study.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.