A number of viral, bacterial and parasitic infections have been proposed to affect the rate of progression of human immunodeficiency virus (HIV) infection [1–6]. Whereas some studies have suggested that infections with hepatitis B virus (HBV) and hepatitis C virus (HCV) [4,6] might be associated with increased mortality in HIV-positive individuals, several research groups have reported a strong association between GB virus C (GBV-C; also called hepatitis G virus) viremia and improved survival in HIV-infected patients [7–10].
GBV-C belongs to the Flaviviridae family [11,12] and, like HIV, it is transmitted parenterally, sexually and vertically. Exposure to GBV-C is equally common in patients who have acquired HIV by parenteral or sexual routes , and it is also prevalent in HIV-negative individuals without risk factors for blood-borne transmission .
GBV-C was initially suspected to be a hepatitis agent. However, most investigations have failed to show GBV-C replication in hepatocytes, or causal associations with hepatitis or other diseases in humans . Ongoing GBV-C infection is diagnosed by detection of viral RNA, whereas resolved infection is followed by the appearance of antibodies against the viral envelope protein E2 (anti-E2). In contrast to the related HCV, GBV-C viremia persists in only a minority of exposed individuals, and there is a strong tendency towards viral clearance over time .
We examined whether the presence of GBV-C RNA at diagnosis of HIV-1 infection can be used to predict disease outcome. Since combination antiretroviral therapy (ART) has a profound effect on survival in HIV infection , we chose to censor study participants at the initiation of such treatment. We had access to serial serum samples from several patients, which allowed us to study the longitudinal course of GBV-C viremia and its association with HIV-1 disease progression.
Materials and methods
Patients with HIV-1 infection registered at the Department of Infectious Diseases at Malmö University Hospital (the only hospital serving the city of Malmö, population 250 000) from January 1984 to September 2000 were eligible for inclusion. Inclusion criteria were availability of a serum sample for GBV-C testing (obtained within 2 years of HIV diagnosis and prior to initiation of ART) and written informed consent for living participants. The samples came from a large serum bank covering more than 30 years (storage temperature −20°C). Out of 365 potential study candidates, 60 patients did not approve inclusion, and 75 patients did not have eligible serum samples stored, leaving a total of 230 subjects who fulfilled the inclusion criteria.
All patients were under the care of physicians specialized in HIV medicine. From 1988 onwards, patients with opportunistic infections or less than 200 × 106 CD4+ cells/l had received prophylaxis for Pneumocystis carinii pneumonia. Since 1996, patients had been offered combination ART according to international guidelines.
Patients were divided into three groups based on their GBV-C status at baseline: GBV-C viremia; anti-E2 antibodies; GBV-C markers absent.
Demographic, clinical, and CD4+ cell-count (determined by flow cytometry) data were obtained from medical records. To define disease stage, clinical criteria according to the 1993 CDC classification system were used; however, CD4+ cell counts were not taken into account for disease classification . In addition, patients with acute HIV infection were considered as a separate group. The patients were followed from the date of HIV diagnosis until one of three endpoints: initiation of dual or triple combination ART, or, for patients who did not receive such therapy, death, or their last visit to our clinic. Four different outcomes were analyzed: all-cause mortality, HIV-related mortality, incidence of AIDS, and incidence of a decline in CD4+ cell counts to below 50% of the baseline level (defined as the mean of the first two counts). For the analysis of the incidence of AIDS, 199 patients were included, and for the analysis of CD4+ cell decline, 142 patients were included (see footnote of Table 3 for exclusion criteria). The decline of CD4+ cell counts was also compared by examining the total and percentage changes per year.
The median age at baseline was 34.7 years, and 89% were men. The median interval between the date of the baseline blood sample used for GBV-C testing and HIV diagnosis was 0 days [interquartile range (IR), 0.00–0.02 days]. The median follow-up time for the whole cohort was 4.3 years (IR, 1.6–8.8 years).
The date of HIV infection was known or could be estimated in 63 patients. In 22 of these, the date was defined by the presence of HIV-1 antigen in the absence of anti-HIV antibodies , and/or by acute retroviral syndrome at diagnosis . In the remaining 41 cases, this date was estimated by calculating the midpoint between the last negative HIV test (within 2 years of diagnosis) and the first positive test. The patients with known and estimated dates of infection were studied as one group and referred to as patients with documented duration of HIV-1 infection.
Follow-up serum samples, obtained as close as possible to the study endpoint, were available in 163 of 230 cases. The median interval between baseline and follow-up samples from these patients was 4.5 years (IR, 2.0–8.2 years). In the survival analysis of patients with longitudinal GBV-C data, the survival time was defined as the interval between HIV diagnosis and the study endpoints described above.
The study was approved by the Research Ethics Committee of Lund University, and was conducted in accordance with the Declaration of Helsinki.
Baseline and follow-up samples were analyzed for GBV-C markers. Baseline samples were also tested for hepatitis B surface antigen (HBsAg), anti-hepatitis B core antibodies (anti-HBc) and anti-HCV antibodies [second and third generation enzyme-linked immunosorbent assay (ELISA)] with standard commercial tests. Positive anti-HCV ELISA results were confirmed by a recombinant immunoblot assay.
The sera were subjected to RNA extraction and reverse transcription, following which a nested polymerase chain reaction was performed to detect GBV-C RNA, using primers directed at the 5′-non-coding region of the GBV-C/HGV genome . Anti-E2 antibodies were detected using the anti-HGenv immunoassay (Hoffmann-La Roche, Basel, Switzerland) according to the manufacturer's instructions.
Tests for GBV-C markers were performed under code without access to clinical data.
Data were analyzed using STATISTICA (version 5.1, 1997; StatSoft, Tulsa, Oklahoma, USA). Baseline characteristics were compared using the χ2 test and Fischer's exact test for categorical variables and the ANOVA test for continuous variables. Analyses of variance were performed across groups. Kaplan–Meier survival curves were employed for time-to-event analysis, using the log-rank test for categorical data and the Cox proportional regression model for continuous data. The Cox proportional regression model was used for multivariate analysis, adjusting for age, HCV infection and hemophilia. P-values lower than 0.05 were considered to be significant.
Markers of GBV-C infection at baseline and follow-up
Patient characteristics at baseline are presented in Table 1. Sixty-two patients had GBV-C viremia (27%), and 69 had anti-E2 (30%). GBV-C RNA and anti-E2 were not detected concomitantly in any patient, thus the total prevalence of GBV-C markers was 57% (131 of 230). Eleven patients showed borderline reactivity in the anti-E2 assay and were considered to be anti-E2-negative.
The distribution of GBV-C markers was similar with respect to age, gender, ethnicity and HBV and HCV co-infection. No patient infected by contaminated blood products exhibited GBV-C viremia, otherwise the prevalence of GBV-C markers was comparable among different transmission categories. GBV-C viremia was significantly less prevalent in patients who presented with AIDS, both when compared across groups (P = 0.008, Table 1), and when compared with all patients with other clinical stages at diagnosis (P = 0.004).
Table 2 presents the patterns of GBV-C markers in 163 patients with an available follow-up sample. Forty-two of the 163 patients (26%) changed their GBV-C status during follow-up.
Mortality and incidence of AIDS
Sixty-two of 78 deaths (79%) during follow-up were attributable to AIDS. Sixty-three of 199 subjects (32%) who did not have AIDS at baseline developed an AIDS-defining condition during follow-up.
Relation with GBV-C status at baseline
GBV-C-viremic patients did not significantly differ from patients lacking GBV-C markers and anti-E2-positive patients when considering all-cause mortality (P = 0.12; Fig.1a), HIV-related mortality (P = 0.18; Fig. 1b), or AIDS incidence (P = 0.84; Fig. 1c).
As we could not exclude an interaction with time during long-term follow-up, we also performed an analysis restricted to the first 8 years of follow-up. This analysis did not demonstrate any significant differences in prognosis according to baseline GBV-C status [P = 0.095 (all-cause mortality), P = 0.14 (HIV-related mortality), P = 0.93 (AIDS incidence)].
Likewise, in the subset of 63 patients with documented duration of HIV infection, all-cause and HIV-related mortality and AIDS incidence were similar in relation to baseline GBV-C status [P-values 0.35 (Fig. 1d), 0.30 and 0.79, respectively].
When the patients were separated into two groups according either to baseline clinical disease stage or CD4+ cell counts, we found that GBV-C viremia was associated with improved survival in the subgroup of patients with symptomatic disease [CDC stages B or C (n = 55); P = 0.01], and in parallel, those with CD4+ cell counts of 300 × 106 cells/l or lower [(n = 53); P = 0.03]. This difference in survival was not observed in patients with asymptomatic infection [(n = 175); P = 0.93], or in those with more than 300 × 106 CD4+ cells/l [(n = 146); P = 0.50].
Relation with GBV-C status at follow-up
Eleven patients with GBV-C RNA at baseline had no detectable viremia at follow-up and did not show detectable anti-E2 seroconversion (Table 2). In comparison with anti-E2-negative subjects with persistent absence, persistent presence or acquisition of GBV-C viremia, the mentioned eleven patients exhibited significantly higher mortality (P = 0.018; Fig. 2), and they were more likely to have developed AIDS (10 of 11 versus 19 of 70 patients; P < 0.001) and to have died of AIDS (8 of 11 versus 16 of 89 patients; P = 0.007) at the end of study. These subjects were characterized by more rapid absolute and percentage decreases in numbers of CD4+ cells (145 versus 56 × 106 cells/l per year, P = 0.006; −0.75 versus −0.25 log/year, P = 0.001). Loss of GBV-C viremia without seroconversion to anti-E2 was not related to age, baseline CD4+ cell counts, or co-infection with HBV or HCV.
By comparison, patients who instead lost anti-E2 during follow-up (Table 2) were older at diagnosis (P = 0.02), but this pattern was not associated with mortality, AIDS incidence, or CD4+ cell decline.
Relation with other baseline characteristics
In univariate analysis, lower age at diagnosis (P < 0.001) and hemophilia (P = 0.03) were associated with decreased all-cause mortality. HIV-related mortality was decreased in patients with lower age at diagnosis (P < 0.001), hemophilia (P = 0.04) and HCV infection (P = 0.01). AIDS-free survival was increased in patients with hemophilia (P < 0.001).
In multivariate analysis, the associations between GBV-C viremia at baseline and the study outcomes remained non-significant [all-cause mortality: relative risk (RR), 0.59; 95% confidence interval (CI), 0.31–1.14; HIV-related mortality: RR, 0.72; 95% CI, 0.35–1.48; AIDS incidence: RR, 0.72; 95% CI, 0.35–1.48]. When using the Cox model, follow-up was restricted to 8 years in order to fulfill the assumption of proportionality.
As patients with hemophilia (who were all HCV-infected) showed improved survival in comparison with other patients, and did not exhibit GBV-C viremia, we performed a separate analysis excluding these subjects to rule out the possibility that they had obscured an association between GBV-C viremia and the studied outcomes. However, the differences in outcomes remained non-significant with regard to baseline GBV-C status (data not shown).
CD4+ cell decline
Assessment of CD4+ cell decline was possible in 142 patients. At baseline, these subjects had a median of 478 × 106 CD4+ cells/l (IR, 280–675 × 106). We observed no significant influence of baseline GBV-C status on CD4+ cell decline (Table 3).
In this HIV cohort, testing for GBV-C markers at diagnosis of HIV-1 infection showed no significant effect of GBV-C viremia on mortality or AIDS incidence during long-term follow-up. Restricting the analysis to patients with documented duration of HIV infection did not affect the results. Our findings indicate that the presence of GBV-C RNA at diagnosis of HIV-1 infection cannot be used to predict a favourable disease course in patients not receiving combination ART.
When designing this study, we hypothesized that the reported associations between GBV-C viremia and improved survival in HIV-1 infection might change if GBV-C RNA disappeared over time in co-infected subjects. In HIV-negative subjects, clearance of GBV-C viremia over time is common [20,21], whereas it seems to occur less frequently in immunocompromized hosts [22,23].
Our results from follow-up in 163 patients show that both GBV-C RNA and anti-E2 can be lost over time in HIV-infected subjects. Fourteen of 44 (32%) patients with GBV-C viremia at baseline had no detectable GBV-C RNA at follow-up. Interestingly, eleven of these 14 subjects did not exhibit seroconversion to anti-E2. Furthermore, these 11 patients had faster disease progression as compared to anti-E2-negative patients with either persistent presence, persistent absence, or acquisition of GBV-C viremia. The absence of detectable anti-E2 seroconversion in patients with loss of GBV-C RNA during follow-up, and the association between this pattern and increased HIV progression, suggests that the mechanism of GBV-C clearance may be different than that observed in HIV-negative subjects [16,21].
When considering a possible interaction between GBV-C and HIV-1, the site of viral replication is clearly of interest. GBV-C replication has been observed in hematopoietic and lymphoid cells, and, in a study by Xiang and coworkers  mainly in CD4+ lymphocytes. If GBV-C and HIV-1 infect the same cell lineages, they may interfere with each other's replication by as yet unknown mechanisms. In accelerating HIV-1 infection, replication of GBV-C might be suppressed, which could explain the inverse correlation between the titers of GBV-C RNA and HIV-1 RNA described by Tillmann et al. . By analyzing samples obtained late in the disease course, when HIV-1 viral load tends to be increased , patients with undetectable GBV-C RNA due to suppressed viral replication could conceivably be misclassified as not having been exposed to GBV-C. Our data demonstrate that absence of GBV-C RNA and anti-E2 in a HIV-positive subject at a certain time point does not necessarily indicate lack of exposure to GBV-C.
Another possible explanation for the rarity of GBV-C viremia in patients with advanced HIV infection could be that a relatively intact immune system might be required for effective GBV-C replication. Lefrère et al.  found a correlation between the viral load of GBV-C and the state of immunocompetence, with the lowest viral loads found in severely immunocompromized hosts.
Our results differ markedly from those reported in most other published studies [8–10]. A probable reason for the discrepancy is that we used serum samples obtained in close association with HIV-1 diagnosis to define GBV-C status. In parallel, the subjects in our study had less advanced disease and higher CD4+ cell counts at inclusion (median 465 × 106 CD4+ cells/l) than subjects reported elsewhere [8–10]. A subgroup analysis of our material showed an association between GBV-C viremia and reduced mortality in patients with symptomatic disease at inclusion and in patients with baseline CD4+ cell counts of 300 × 106 cells/l or lower. The lower mortality associated with GBV-C viremia in patients with advanced HIV infection is compatible with the hypothesis that the GBV-C-viremic subjects had lower HIV-1 viral loads [9,27] and therefore also slower clinical progression compared to patients at similar disease stage without GBV-C viremia.
Since we censored follow-up at the time of initiation of combination ART, bias might have been introduced concerning the reasons for initiating ART in individual patients, and, more importantly, concerning whether patients were included before or after such therapy became available. In order to exclude the latter form of bias, we analyzed patients recruited before 1996 (with follow-up endpoint 30 June 1996) separately, and the results were similar to those described above (data not shown). We found no differences in indications for starting ART associated with GBV-C status, as the proportions of patients who had initiation of ART as their study endpoint were similar with regard to baseline GBV-C markers (Table 1).
Among base-line characteristics other than GBV-C status, we found that HCV co-infection was associated with reduced HIV-related mortality. In comparison with investigations examining the interaction between HIV-1 and HCV [4,28,29], few of our patients (18 of 230; 7.8%) were infected by intravenous drug use. When patients with hemophilia were excluded from the analysis, the association between HCV infection and HIV-related mortality lost statistical significance (P = 0.07). Our results do not support a negative influence of HCV in HIV infection, as reported by other authors [4,29]. However, we could exclude confounding by HCV co-infection on the effects of GBV-C on the course of HIV-1 infection.
In conclusion, our findings suggest that GBV-C viremia in HIV-1 infected individuals may represent a phenomenon secondary to the rate of HIV-1 progression rather than an independent prognostic factor. For the evaluation of patients newly diagnosed with HIV infection, GBV-C RNA testing cannot be used to predict the natural disease course. However, an interaction between the two viruses appears to exist, since loss of GBV-C viremia without detectable anti-E2 seroconversion was associated with increased HIV progression, and this merits further investigation.
We are indebted to Jan-Åke Nilsson for help with statistical analyses, Prof. Joakim Dillner for valuable advice, and Dr. Torkil Moestrup for providing clinical data from the beginning of the HIV epidemic.
Sponsorship: Supported by grants from the Medical Faculty of Lund University, the Cancer Foundation of Malmö University Hospital, the Alfred Österlund Foundation, and the Donation Funds of Malmö University Hospital.
1. Badri M, Ehrlich R, Wood R, Pulerwitz T, Maartens G. Association between tuberculosis and HIV disease progression in a high tuberculosis prevalence area. Int J Tuberc Lung Dis
2. Watt G, Kantipong P, de Souza M, Chanbancherd P, Jongsakul K, Ruangweerayud R, et al
. HIV-1 suppression during acute scrub-typhus infection. Lancet
3. Christensen JK, Eugen-Olsen J, Sörensen M, Ullum H, Gjedde SB, Pedersen BK, et al
. Prevalence and prognostic signficance of infection with TT virus in patients infected with human immunodeficiency virus. J Infect Dis
4. Greub G, Ledergerber B, Battegay M, Grob P, Perrin L, Furrer H, et al
. Clinical progression, survival, and immune recovery during antiretroviral therapy in patients with HIV-1 and hepatitis C virus coinfection: the Swiss Cohort Study. Lancet
5. Ockenga J, Tillmann HL, Trautwein C, Stoll M, Manns MP, Schmidt RE. Hepatitis B and C in HIV-infected patients. J Hepatol
6. Hoffman IF, Jere CS, Taylor TE, Munthali P, Dyer JR, Wirima JJ, et al
. The effect of Plasmodium falciparum malaria on HIV-1 RNA blood plasma concentration. AIDS
7. Lefrère J-J, Roudot-Thoraval F, Morand-Joubert L, Petit JC, Lerable J, Thauvin M, et al
. Carriage of GB virus C/hepatitis G virus RNA is associated with a slower immunologic, virologic, and clinical progression of human immunodeficiency virus disease in coinfected persons. J Infect Dis
8. Xiang J, Wünschmann S, Diekema DJ, Klinzman D, Patrick KD, George SL, et al
. Effect of coinfection with GB virus C on survival among patients with HIV infection. N Engl J Med
9. Tillmann HL, Heiken H, Knapik-Botor A, Heringlake S, Ockenga J, Wilber JC, et al
. Infection with GB virus C and reduced mortality among HIV-infected patients. N Engl J Med
10. Yeo AET, Matsumoto A, Hisada M, Shih JW, Alter HJ, Goedert JJ. Effect of hepatitis G virus infection on progression of HIV infection in patients with hemophilia. Ann Intern Med
11. Simons JN, Leary TP, Dawson GJ, Pilot-Matias TJ, Muerhoff AS, Schlauder GG, et al
. Isolation of novel virus-like sequences associated with human hepatitis. Nat Med
12. Linnen J, Wages J Jr., Zhang-Keck Z-Y, Fry KE, Krawczynski KZ, Alter H, et al
. Molecular cloning and disease association of hepatitis G virus: a transfusion-transmissible agent. Science
13. Bourlet T, Guglielminotti C, Evrard M, Berthelot P, Grattard F, Fresard A, et al
. Prevalence of GBV-C/hepatitis G virus RNA and E2 antibody among subjects infected with human immunodeficiency virus type 1 after parenteral and sexual exposure. J Med Virol
14. Björkman P, Nauclér A, Winqvist N, Mushahwar I, Widell A. A case-control study of transmission routes for GB virus C/hepatitis G virus in Swedish blood donors lacking markers for hepatitis C virus infection. Vox Sang
15. Bowden S. New hepatitis viruses: contenders and pretenders. J Gastroenterol Hepatol
16. Thomas DL, Vlahov D, Alter HJ, Hunt JC, Marshall R, Astemborski J, et al
. Association of antibody to GB virus C (hepatitis G virus) with viral clearance and protection from reinfection. J Infect Dis
17. Palella FJ Jr, Delaney KM, Moorman AC, Loveless MO, Fuhrer J, Satten GA, et al
. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. N Engl J Med
18. Centers for Disease Control and Prevention. 1993 revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. MMWR
19. Björkman P, Sundström G, Widell A. Hepatitis C virus and GB virus C/hepatitis G virus viremia in Swedish blood donors with different alanine aminotransferase levels. Transfusion
20. Wang J-T, Tsai F-C, Lee C-Z, Chen P-J, Sheu J-C, Wang T-H, et al
. A prospective study of transfusion-transmitted GB virus C infection: similar frequency but different clinical presentation compared with hepatitis C virus. Blood
21. Björkman P, Sundström G, Veress B, Widell A. Assessment of liver disease and biochemical and immunological markers in Swedish blood donors with isolated GB virus C/hepatitis G virus viremia. Vox Sang
22. Goubau P, Liu HF, Goderniaux E, Bourtonboy G. Influence of CD4+lymphocyte counts on GB virus C/hepatitis G virus carriership in HIV-positive individuals. J Med Virol
23. Devereux H, Sabin CA, Kinson Z, Brown D, Griffioen A, Dusheiko GM, et al
. Influence of HIV-1 infection on GBV-C infection in multiply infected haemophilic patients. J Med Virol
24. Xiang J, Wünschmann S, Schmidt W, Shao J, Stapleton JT. Full-length GB virus C (hepatitis G virus) RNA transcripts are infectious in primary CD4-positive T cells. J Virol
25. Holodniy M, Katzenstein DA, Sengupta S, Wang AM, Casipit C, Schwartz DH, et al
. Detection and quantification of human immunodeficiency virus RNA in patient serum by use of the polymerase chain reaction. J Infect Dis
26. Lefrère J-J, Férec C, Roudot-Thoraval F, Loiseau P, Cantaloube JF, Biagini P, et al
. GBV-C/hepatitis G virus (HGV) RNA load in immunodeficient and in immunocompetent individuals. J Med Virol
27. Brumme ZL, Chan KJ, Dong WWY, Mo T, Wynhoven B, Hogg RS, et al
. No association between GB virus-C viremia and virological or immunological failure after starting initial antiretroviral therapy. AIDS
28. Sulkowski MS, Moore RD, Mehta SH, Chaisson RE, Thomas DL. Hepatitis C and progression of HIV disease. JAMA
29. Piroth L, Duong M, Quantin C, Abrahamowicz M, Michardiere R, Aho LS, et al
. Does hepatitis C virus co-infection accelerate clinical and immunological evolution of HIV-infected patients? AIDS