Herpesvirus and GB Virus C Infections in Men With and Without HIV-1 Who Have Sex with Men

Abstract

In men with (MWH) and without (MWOH) HIV-1 infection, we longitudinally evaluated the prevalence of human herpesvirus (HHV) and GB virus C (GBV-C) infections and examined associations with plasma HIV-1 load and inflammatory markers. We analyzed 11,874 plasma samples (collected from 1984 to 2009) from 1,882 men who have sex with men in the Multicenter AIDS Cohort Study to quantify four HHVs [cytomegalovirus (CMV), Epstein-Barr virus (EBV), human herpesvirus 8 (HHV-8), and human herpesvirus 6 (HHV-6)] and GBV-C. HHV and GBV-C viral loads were measured using quantitative Polymerase Chain Reaction (PCR) and Reverse Transcription (RT)-PCR, respectively. Differences in viral prevalence and concentrations were assessed using multivariable logistic and linear regression. Associations between HHV viremia and 24 inflammatory and immune activation biomarkers were evaluated using generalized gamma models. Age-related biomarker trajectories were analyzed using linear mixed models among MWH with sustained suppression due to HIV-1 antiretroviral therapy. MWH with detectable plasma HIV-1 RNA had significantly higher odds of CMV, EBV, and HHV-8 viremia compared with those with undetectable plasma HIV-1 RNA, who in turn had higher odds than MWOH. CMV and EBV viremia were associated with higher levels of multiple inflammatory markers. Among MWH with consistent viral suppression, no significant differences were observed in age-related biomarker trajectories. Overall, active HHV infection, as indicated by viremia, was associated with significantly higher plasma HIV-1 loads and increased inflammatory marker levels—particularly in the presence of detectable plasma HIV-1 RNA. Our study supports the hypothesis that active HHV infections may exacerbate HIV-1 disease progression in MWH.

Keywords

HIV-1 CMV EBV GBV-C inflammation reactivation

Introduction

Human herpesviruses (HHV) and human immunodeficiency virus type 1 (HIV-1) cause persistent, lifelong infections. HHV [i.e., herpes simplex virus (HSV)-1, HSV-2, varicella zoster virus, Epstein-Barr virus (EBV), cytomegalovirus (CMV), HHV-6, HHV-7, and HHV-8] are widespread and infect a significant portion of the global population. Infection rates vary depending on the specific HHV, the population studied, and geographic factors, with the most prevalent being EBV, CMV, HHV-6 and HHV-7.¹ It is estimated that nearly 100% of men who have sex with men (MSM) and who are infected with HIV-1 (MWH) are infected with EBV, and 91%–94% are infected with CMV.² HHV establishes latent infection in host cells specific for each type of virus and can periodically reactivate during stress, trauma, infection, and immune suppression.^3,4 Following primary infection, CMV remains latent in monocytes, dendritic cells, and endothelial cells.⁵ CMV reactivation often occurs in immunocompromised individuals, including MWH, and can lead to CMV retinitis, pneumonia, encephalitis, and nephritis. Similarly, after primary infection, EBV becomes latent in B lymphocytes and epithelial cells.⁶ Virus reactivation can lead to lymphoproliferative disorders and cancer development, such as nasopharyngeal carcinoma, Hodgkin’s lymphoma, and Burkitt’s lymphoma, especially in immunocompromised individuals with HIV-1/AIDS and organ transplant recipients.

GB virus C (GBV-C) is a nonpathogenic human RNA pegivirus, which can be transmitted parenterally, sexually, and vertically. GBV-C is highly prevalent among MSM, especially in people with HIV-1 (PWH).^7–10 GBV-C infection is usually benign and asymptomatic, with no significant health impact on its own. However, GBV-C infection has a potential interaction with other viral infections, especially HIV-1.^7,9 Although some studies of PWH have noted a survival benefit of coinfection with GBV-C,^7–9 other studies have not found this effect.^10,11

MWH and men without HIV-1 (MWOH) often have multiple chronic viral infections. Since each virus elicits different immune responses, co-reactivation may have a stronger or weaker association with inflammation and HIV-1 disease progression than any single virus alone. However, limited data exist on reactivation of multiple herpesviruses simultaneously in the context of serological markers in PWH.¹² Likewise, only limited longitudinal data exist on the prevalence of HHV reactivation and GBV-C viremia in the context of HIV-1 infection.^13–16 We report here on the prevalence and quantity of EBV, CMV, HHV-6, HHV-8, and GBV-C in stored plasma from MWH and MWOH over multiple time points from 1984 to 2009. We tested for associations between HIV-1 status (HIV-1 negative, HIV-1 positive with detectable plasma HIV-1 RNA, and HIV-1 positive with undetectable plasma HIV-1 RNA) and the presence/quantity of these five viruses. To better understand which inflammatory pathways are relevant during viral reactivation, we compared serum concentrations of 24 inflammatory biomarkers of study participants with HHV and GBV-C viremia from the same time points. Since herpesvirus reactivation often occurs in elderly persons due to age-associated decline of immune control,¹⁷ we also conducted a longitudinal analysis among men with consistently suppressed HIV-1 infection to test whether age-related trajectories of seven selected inflammatory biomarkers differed between those with and without HHV and GBV-C viremia.

Materials and Methods

Study participants and plasma samples

The Multicenter AIDS Cohort Study (MACS) was a prospective cohort study of HIV-1 infection of MSM at four sites in the USA (Johns Hopkins University, Baltimore, MD/Washington, DC; Northwestern University, Chicago, IL; University of California, Los Angeles, CA; and University of Pittsburgh, Pittsburgh, PA). Enrollment began in April 1984, with later waves of recruitment; details of the cohort have been described elsewhere.^18,19 Participants were assessed at semiannual study visits with standardized interviews, physical examinations, and phlebotomy for concurrent laboratory testing and storage of heparinized plasma and serum (at −80°C) and viable peripheral blood mononuclear cells (at −135°C).

Biological samples were the same as those selected for a separate study of biomarkers of inflammation and immune activation, described previously.²⁰ Briefly, samples tested were from MACS participants who either (1) had a known date of HIV-1 seroconversion (selected samples were from visits both before and after seroconversion), or (2) had HIV-1 infection at study entry and had initiated antiretroviral therapy (ART) (selected samples were from visits before and after ART initiation). ART during this period included various combinations of nucleoside reverse transcriptase inhibitors, protease inhibitors, non-nucleoside reverse transcriptase inhibitors, and various combination HIV-1/AIDS treatments.^21,22 In addition, multiple samples were selected from MWOH, including all such men in the MACS with active hepatitis C virus (HCV) infection.

Serologic markers of inflammation and immune activation were quantified as previously described using Meso Scale Discovery (MSD, Gaithersburg, MD, USA) and Luminex (Luminex, Austin, TX, USA) multiplex platforms, with the exception of C-reactive protein, which was measured using an immunonephelometric assay by a clinical reference laboratory (Quest Diagnostics, Madison, NJ, USA).²⁰ CD4⁺ T cell counts were measured with flow cytometry.²³ Plasma HIV-1 RNA levels were primarily measured with either the Roche Cobas-Taqman v2.0 assay, sensitive to ≥20 copies of HIV-1 RNA/mL plasma, or the Roche ultrasensitive assay sensitive to ≥50 copies. A small number of older samples were measured with the Chiron 2nd generation assay sensitive to ≥500 copies or the Roche 2nd generation assay sensitive to ≥400 copies. For the 2.7% of samples (n = 281) from MWH with missing HIV-1 viral load, HIV-1 detectability was assigned based on ART exposure: samples collected prior to ART use were classified as detectable HIV-1 RNA (n = 269), while samples collected during reported ART use were classified as undetectable HIV-1 RNA (n = 12). We also tested the robustness of our findings to this assumption by performing each analysis with these samples omitted. We classified samples as HIV-1 negative (HIV-1 negative), HIV-1 positive with detectable plasma HIV-1 RNA (HIV-1 positive detectable), and HIV-1-positive with undetectable plasma HIV-1 RNA (HIV-1 positive undetectable).

Nucleic acid extraction and quantification of HHV and GBV-C

Whole nucleic acids were extracted from the heparinized plasma samples as described elsewhere.^24,25 Whole nucleic acid was isolated by an NucliSENS EasyMAG (bioMérieux; Boston, MA) automated extraction system from 1 mL plasma containing a fixed amount of RNA internal control virus [equine arteritis virus (EAV)]²⁶ and a fixed amount of DNA internal control virus (phocine herpesvirus, PhHV).²⁷ Quantities of EBV, CMV, HHV-6, and HHV-8 viral DNA and GBV-C RNA were measured as described previously.^24,25,28 Quantitative real-time Polymerase Chain Reaction (PCR) targeting EBV, CMV, HHV-6, HHV-8, and PhHV, and quantitative Reverse Transcription (RT) real-time PCR targeting GBV-C and EAV, were assessed along with the extracted nucleic acids. To ensure the efficiency of nuclear acid extraction, RT, and real-time PCR, the nucleic acids were re-extracted followed by PCR or RT-PCR for any samples with a quantity of PhHV or EAV below the cutoff values.

Statistical analysis

Covariates for analysis were chosen a priori as possible confounders of the relationships between exposures and outcomes; these were time-updated where appropriate. HIV-1-specific covariates were defined at the exact date of viral coinfection sampling. Other covariates were taken from the exact date where possible or were carried forward from the most recent value within 2 years of the sample. Hepatitis B (HBV) infection was defined by presence of hepatitis B surface antigen; HCV infection was defined by detectable HCV RNA. Smoking (yes/no), heavy drinking (>13 alcoholic drinks/week), and injection drug use (yes/no) were defined by self-report. Body mass index (kg/m2) and age were treated as continuous. Uncontrolled diabetes was defined as hemoglobin A1C ≥ 6.5% or fasting glucose ≥126 mg/dl. Anemia was defined as <5th percentile of hemoglobin among the general population by age or taking erythropoietin.

Because of the high prevalence of these viral coinfections in the source population, it is likely that any detectable viremia represents a reactivation event rather than incident infection. However, some participants may have been uninfected with a given virus at baseline. To improve the validity of our assumption of reactivation, we tested the plasma samples for viral antibodies. To this end, available baseline plasma samples without detectable DNA for one of these herpesviruses (n = 1,008) were measured by commercially available Enzyme-Linked Immunosorbent Assay kits for the presence of IgG antibodies to CMV (GenWay, San Diego, CA), EBV (Abcam, Cambridge, UK), and HHV-6 (MyBioSource, San Diego, CA) following the manufacturer’s protocols. We then tested all available subsequent study samples for men whose baseline samples were negative for viral antibodies [54 CMV baseline negative, 38 EBV baseline negative, 77 HHV-6 baseline negative], including a retest of the baseline sample. Men testing negative for antibodies for a given virus at baseline and without detectable viral DNA were classified as seronegative, and their person-visits up to any first evidence of infection were excluded from virus-specific reactivation analyses. For all but 2 men in this study, testing for antibodies to HHV-8 (defined as antibodies against either the latency-associated nuclear antigen or the lytic antigen encoded by orfK8.1) had been performed on a subset of visits (n = 4,383) for a previous analysis.²⁹ Participants were classified as seronegative at baseline and excluded from HHV-8-specific analyses if they tested HHV-8 antibody-negative for at least one visit and did not test HHV-8 antibody-positive or HHV-8 DNA-positive until after the baseline study visit. We did not perform antibody testing for GBV-C. Supplementary Figure S1 displays a schematic showing antibody testing and classification of prior HHV exposure.

Estimates of the prevalence of viremia, along with unadjusted and adjusted differences in proportion by HIV-1 status, were assessed via logistic regression adjusted for repeated measures per individual using generalized estimating equations. Differences in viral DNA/RNA concentrations by HIV-1 group were assessed via multivariable linear regression on the natural log of viral loads (ln copies/mL), adjusted for repeated measurements.

We compared the serum concentrations of 24 biomarkers of inflammation and immune activation, measured for a prior study,³⁰ between samples with and without active HHV/GBV-C viremia, stratified by HIV-1 positivity. To do so, we fit biomarker-specific multivariable generalized gamma models, holding scale and shape parameters constant but allowing the location parameter to vary. We adjusted for multiple testing by controlling the false discovery rate (FDR) at 0.05 per set of 24 markers tested.

For a selection of seven biomarkers that (a) showed differences by herpesvirus/GBV-C viremia and (b) had previously observed associations with morbidity and mortality in PWH, we fit linear mixed models to test whether long-term trajectories of these biomarker concentrations differed between men with and without CMV and EBV reactivation.

For this analysis, we restricted the study population to MWH meeting a certain threshold for consistent viral suppression, defined by at least two visits while taking ART with at least 75% of samples from these visits having undetectable plasma HIV-1 RNA. Presence or absence of CMV and EBV reactivation was defined during the window of consistent HIV-1 viral suppression. Biomarker values below the lower limit of detection were imputed using a lognormal distribution. We adjusted for multiple tests by controlling the FDR at 0.05 for each set of 14 parameters tested (7 biomarkers, slope, and intercept) for each virus.

Results

Characteristics of study population

We included 11,874 person-visits contributed by 1,882 MACS participants from 1984 to 2009 in this study. Each participant contributed a median of 4 (Interquartile range [IQR]: 4, 8; range 1–29) visits to the study; the median time between visits for each participant was 1.1 years (IQR: 0.8, 2.0; range 0.1–20.7). Participants could contribute samples to more than one HIV-1 category (HIV-1 negative, HIV-1 positive detectable, HIV-1 positive undetectable); thus, 819 contributed to only one HIV-1 category, while 144 contributed to all three categories. Characteristics of the study population and the person-visits included in the study are displayed in Table 1.

Table 1.

Characteristics of Study Population and Person-Visits

	HIV-1 negative	HIV-1 positive detectable	HIV-1 positive undetectable	Total
Individuals at baseline	590	944	348	1882
Age in years at baseline	37 (30,43)	40 (35,46)	40 (35,46)	39 (34,45)
% Asian/Pacific/Nat. Amer.	2%	4%	10%	4%
% Black, Hispanic	0%	1%	1%	1%
% Black, non-Hispanic	23%	27%	33%	27%
% White, Hispanic	5%	5%	10%	6%
% White, non-Hispanic	68%	60%	41%	59%
% Other race	1%	3%	4%	3%
Person-visits	1441	6460	3973	11,874
Age in years at visit	43 (35,50)	41 (36,48)	48 (42,54)	44 (37,50)
Current smoking	40%	34%	28%	33%
Heavy drinking	11%	9%	5%	8%
Injection drug use	5%	2%	2%	3%
Hepatitis B infection	3%	4%	4%	4%
Hepatitis C infection	17%	7%	8%	9%
Diabetes	7%	4%	9%	6%
Anemia	10%	21%	13%	17%
BMI (kg/m²)	24.8 (22.7,27.8)	24.2 (22.2,26.5)	24.6 (22.5,27.1)	24.4 (22.4,26.9)
CD4⁺ T cell count/μL	—	455 (287,662)	587 (416,789)	506 (332,720)
CD4 ≤200	—	15%	4%	11%
CD4 201–350	—	19%	12%	17%
CD4 351–500	—	22%	20%	22%
CD4 >500	—	43%	63%	51%
HIV-1 viral load (copies/mL)	—	14,422 (1942,59,209)	—	901 (40,24,594)
ART use at visit tested	—	26%	84%	49%
Years since seroconversion	—	3.5 (1.4, 7.5)	12.7 (7.2, 17.2)	5.1 (1.7, 11.0)
Years since ART initiation	—	4.3 (1.9, 7.0)	6.0 (3.0, 8.9)	5.3 (2.3, 8.2)

ART, antiretroviral therapy; BMI, Body Mass Index.

Virus antibody results

To assess the prevalence of CMV, EBV, and/or HHV-6 infection in the study participants at beginning of the study, antibody testing for CMV, EBV, and/or HHV-6 was performed on baseline samples among men with undetectable viral DNA. If any HHV antibody was negative at baseline, all subsequent samples of the same participant were tested for the antibody as well (Supplementary Fig. S1). Results of testing for antibodies against CMV, EBV, HHV-8, and HHV-6 are shown in Supplementary Table S1. For analyses of viral reactivation, we retained 11,442 samples for CMV analyses, 11,712 samples for EBV analyses, 11,352 samples for HHV-6 analyses, 7,270 samples for HHV-8 analyses, and 11,791 samples for GBV-C analyses. We also identified men whom we suspect to have acquired primary herpesvirus infection during follow-up (CMV: n = 10, EBV: n = 2, HHV-6: n = 1, HHV-8: n = 227), and men who were persistently seronegative without any detectable viremia (CMV: n = 42, EBV: n = 1, HHV-6: n = 71, HHV-8: n = 515) (Supplementary Fig. S1, Supplementary Table S1).

Prevalence of HHV and GBV-C viremia

Over the course of the study in the full population (including those seronegative at baseline for antibodies to one or more of the viruses), 331 men (18%) never tested positive for any viral DNA/RNA, while 595 men (32%) tested positive for only one virus, and 956 men (51%) tested positive for more than one. In 4,414 plasma samples (38%), only one virus was detected; in 2,054 samples (17%), more than one of the viruses were detected, with four of the five viruses detected in 32 samples (0.3%). After excluding men with evidence of no prior infection, the proportion of men who ever had detectable reactivated viremia ranged from 4% for HHV-6 up to 60% for EBV; 13% of participants had 4 or more samples with detectable EBV DNA.

The prevalence of reactivated viruses in the study population from repeated measures logistic regression, stratified by HIV-1 status, is shown in Table 2. Across all groups, the prevalence of detectable DNA/RNA was 11.8% for CMV, 25.2% for EBV, 1.2% for HHV-6, 10.6% for HHV-8, and 29.4% for GBV-C. Prevalence of viremia for CMV, EBV and HHV-8 was higher in both HIV-1 positive groups relative to the HIV-1 negative group (p < .05 for all comparisons). For all viruses, the prevalence of reactivated viruses was higher in HIV-1 positive detectable samples relative to HIV-1 positive undetectable samples (p < .05).

Table 2.

Estimated Prevalence of HHV/GBV-C Reactivated Viremia across HIV-1 Groups

	n	Prevalence	95% LL	95% UL	p1	p2
CMV
All HIV-1 groups	11,442	11.80%	11.00%	12.70%	—	—
HIV-1 negative	1,288	3.40%	2.50%	4.70%	—	—
HIV-1 positive detectable	6,294	16.10%	15.00%	17.40%	<.005	—
HIV-1 positive undetectable	3,860	8.30%	7.40%	9.30%	<.005	<.005
EBV
All HIV-1 groups	11,712	25.20%	24.00%	26.50%	—	—
HIV-1 negative	1,423	10.60%	9.10%	12.30%	—	—
HIV-1 positive detectable	6,362	31.80%	30.10%	33.50%	<.005	—
HIV-1 positive undetectable	3,927	22.20%	20.50%	24.00%	<.005	<.005
HHV-8
All HIV-1 groups	7,270	10.60%	9.50%	11.70%	—	—
HIV-1 negative	420	3.20%	1.70%	6.00%	—	—
HIV-1 positive detectable	4,182	13.80%	12.30%	15.40%	<.005	—
HIV-1 positive undetectable	2,668	7.20%	6.10%	8.60%	.013	<.005
HHV-6
All HIV-1 groups	11,352	1.20%	0.90%	1.60%	—	—
HIV-1 negative	1,258	1.60%	0.80%	3.00%	—	—
HIV-1 positive detectable	6,263	1.30%	0.90%	1.80%	.541	—
HIV-1 positive undetectable	3,831	0.80%	0.50%	1.20%	.051	.015
GBV-C
All HIV-1 groups	11,791	29.40%	27.60%	31.30%	—	—
HIV-1 negative	1,434	33.20%	30.20%	36.40%	—	—
HIV-1 positive detectable	6,409	31.20%	29.10%	33.40%	.158	—
HIV-1 positive undetectable	3,948	25.40%	23.10%	27.70%	<.005	<.005

Notes: Prevalence estimates were from logistic regression adjusted for repeated measures. Virus-specific estimates exclude person-visits before and during suspected primary infection.

p¹: p value for comparison with HIV-1 negative group.

p²: p value for comparison with HIV-1 positive detectable group.

CMV, cytomegalovirus; HHV, human herpesvirus; GBV-C, GB virus C; LL, Lower Limit; UL, Upper Limit.

Estimated odds ratios (ORs) from multivariable repeated measures logistic regression models, estimating the relative likelihood of observing viral reactivation at a given study visit, are presented in Table 3. For CMV, EBV, and HHV-8, both groups with HIV-1 had higher prevalence odds (p < .05) than HIV-1 negative group, with HIV-1 positive detectable group having the highest ORs for each virus [CMV: 4.76 (3.38, 6.7); EBV: 3.91 (3.23, 4.73); HHV-8: 5.01 (2.55, 9.85)]. Among MWH, higher HIV-1 viral load was associated (p < .05) with higher prevalence of all viruses except GBV-C (ORs are displayed in Table 3). Lower CD4⁺ T cell counts were associated (p < .05) with higher prevalence of CMV, EBV, and HHV-8.

Table 3.

Adjusted Odds Ratios for Prevalence of HHV/GBV-C Reactivated Viremia

	HIV-1 negative (ref.)	HIV-1 positive detectable	HIV-1 positive undetectable	HIV-1 RNA (log copies/mL)	CD4⁺ cell count (100 cells/ul)
	HIV-1 negative (ref.)	HIV-1 positive detectable	HIV-1 positive undetectable	HIV-1 RNA (log copies/mL)	CD4⁺ cell count (100 cells/ul)	CMV	—	4.76 (3.38, 6.70)^*	2.42 (1.70, 3.46)^*	1.09 (1.06, 1.12)^*	0.84 (0.81, 0.87)^*
EBV	—	3.91 (3.23, 4.73)^*	2.29 (1.85, 2.83)^*	1.07 (1.05, 1.09)^*	0.95 (0.93, 0.97)^*
HHV-8	—	5.01 (2.55, 9.85)^*	2.47 (1.23, 4.97)^*	1.12 (1.09, 1.15)^*	0.96 (0.92, 0.99)^*
HHV-6	—	0.90 (0.47, 1.74)	0.52 (0.27, 1.03)	1.09 (1.02, 1.17)^*	1.03 (0.95, 1.11)
GBV-C	—	1.14 (0.99, 1.30)	1.12 (0.95, 1.31)	0.99 (0.98, 1.01)	1.00 (0.98, 1.02)

Note: Models were adjusted for age, body mass index, current smoking status, alcohol consumption, injection drug use, active hepatitis B infection, active hepatitis C infection, diabetes, and anemia. Virus-specific estimates exclude person-visits before and during suspected primary infection.

*Indicates p < .05.

EBV, Epstein-Barr virus; UL, Upper Limit.

Viral quantitation

Distributional plots of log viral loads for each virus, by HIV-1 category, are displayed in Figure 1. There were no significant differences in the viral loads of these viruses measured among HIV-1 positive detectable, HIV-1 positive undetectable, and HIV-1 negative groups. The medians and interquartile ranges are described in Supplementary Table S2a. HIV-1 positive detectable samples exhibited the greatest range in CMV, EBV, HHV-8, and HHV-6 viral loads, followed by HIV-1 positive undetectable samples. For the small number (n = 65) of suspected primary HHV infections, log viral load distributions are displayed in Supplementary Table S2b.

FIG. 1.

Viral concentrations by HIV-1 Group. Violin plots display range and density for each viral load. A: HIV-1 negative; B: HIV-1 positive virus detectable; C: HIV-1 positive virus undetectable. Black diamonds indicate 25th, 50th, and 75th percentiles log10 copies virus/mL; yellow diamond indicates mean log10 copies virus/mL. Quartiles are indicated with blue shading. Suspected primary infections were excluded.

Results from multivariable linear regression models fit to the natural log of viral loads are displayed in Table 4. For CMV, HIV-1 status was associated with the magnitude of viremia: HIV-1 positive detectable samples had higher quantities of CMV DNA compared with HIV-1 negative samples (0.91 in copies/mL higher, 95% Confidence Interval [CI]: 0.18–1.64), with a similar result for HIV-1 positive undetectable samples (0.98 in copies/mL higher, 95% CI: 0.12, 1.84). HIV-1 positive undetectable samples also had higher quantities of GBV-C RNA relative to HIV-1 negative samples (0.61 in copies/mL higher, 95% CI: 0.09–1.12). However, HIV-1 status did not predict the magnitude of detectable viremia for the other three viruses (EBV, HHV-8, and HHV-6).

Table 4.

Estimated Differences in the Natural Log of HHV/GBV-C Viral Loads (In Copies/Ml) by HIV-1 Group

	HIV-1 negative (ref.)	HIV-1 positive detectable	HIV-1 positive undetectable
CMV	—	0.91 (0.18, 1.64)*	0.98 (0.12, 1.84)*
EBV	—	0.01 (−0.29, 0.31)	−0.16 (−0.46, 0.13)
HHV-8	—	0.39 (−0.55, 1.32)	0.26 (−0.70, 1.21)
HHV-6	—	0.20 (−2.60, 2.99)	−1.64 (−5.23, 1.95)
GBV-C	—	−0.04 (−0.50, 0.42)	0.61 (0.09, 1.12)*

Note: The data were analyzed by linear regression models adjusted for age, body mass index, current smoking status, alcohol consumption, injection drug use, active hepatitis B infection, active hepatitis C infection, diabetes, and anemia. Estimates exclude suspected primary infections.

*Indicates p < .05.

Inflammatory biomarker concentrations

Figure 2 displays results from multivariable generalized gamma models, stratified by HIV-1 positivity, comparing concentrations of 24 markers of inflammation and immune activation between samples with and without HHV/GBV-C viremia. After adjustment for multiple tests, statistically significant higher concentrations were observed among CMV-positive samples for 15 out of the 24 markers, but only among those samples that were also HIV-1 positive. The largest estimated percentage difference was for C-reactive protein (CRP; 32%, 95% CI: 19%–44%). Among HIV-1 negative samples, there were no statistically significant biomarker differences by CMV viremia. For EBV, the pattern was substantially different: in both HIV-1 positive and HIV-1 negative samples, 11 markers had statistically significant higher concentrations among samples with EBV viremia relative to those without EBV viremia, and 2 markers were higher only among HIV-1 negative samples. Of these 13 markers, 11 were among those that also exhibited higher concentrations among HIV-1 positive samples with CMV viremia: CRP, IFN-γ, CXCL10, IL-2, sCD27, IL-6, TNF-α, SIL-2Rα, IL-10, sTNFR2, and sCD14. There were few associations between viremia and inflammatory biomarker concentrations for HHV-6 (N = 3), HHV-8 (N = 2), and GBV-C (N = 3), regardless of HIV-1 status.

FIG. 2.

Differences in plasma biomarker concentrations by presence/absence of CMV, EBV, HHV-8, HHV-6, and GBV-C viremia. Data are from generalized gamma models adjusted for age, body mass index, current smoking status, alcohol consumption, injection drug use, active hepatitis B infection, active hepatitis C infection, diabetes, and anemia. Squares: HIV-1 positive samples; Circles: HIV-1 negative samples. Bars represent 95% confidence intervals. Red fill indicates statistical significance after controlling false discovery rate at 0.05. CMV, cytomegalovirus; EBV, Epstein-Barr virus; HHV, human herpesvirus; GBV, GB virus C.

Biomarker trajectories among men with consistent HIV-1 suppression, with and without herpesvirus reactivation

We chose seven biomarkers that were associated with CMV and/or EBV reactivation in the preceding analysis and that had previously demonstrated associations with all-cause mortality:³⁰ CRP, IL-6, sIL-2Rα, CXCL13, sCD14, CXCL10, and TNF-α. We selected a subpopulation of 678 men who demonstrated consistent HIV-1 suppression over a total of 3,050 visits. During this period of consistent suppression, 149 men experienced CMV reactivation and 324 experienced EBV reactivation. Visual inspection of raw individual trajectories showed no clear differences by EBV or CMV reactivation status.

The results from multivariable linear mixed models, with the natural log of biomarker concentrations modeled as a linear function of age, are displayed in Supplementary Table S3a and 3b. Among those with CMV reactivation (Supplementary Table S3a), there was evidence at α = 0.05 for higher concentrations at age 45 (intercept) of CRP, CXCL10, and TNF-α; and lower concentrations of sCD14, relative to those without CMV reactivation. The rate of change of log CXCL10 concentrations per 10 years of age (slope) was lower among those with CMV reactivation than among those without (−0.04 vs. 0.08, p = .03). Among those with EBV reactivation (Supplementary Table S3b), the slope for IL-6 was nearly double that of those without EBV reactivation (0.24 vs. 0.12, p = .03). However, after adjustment for multiple tests, the only parameter that retained statistical significance was the higher CRP intercept for those with CMV reactivation (−0.56 vs. −0.83, p = .0024).

Discussion

HHV viremia in people with preexisting anti-HHV serum antibodies is a definitive marker of reactivated HHV infection. Periodic CMV, EBV, and/or HHV8 reactivation has a profound impact on local and systemic immune systems, leading to chronic immune activation. Compared with other reactivating HHVs causing milder or more localized effects, CMV and EBV reactivation often cause significant clinical disease, particularly in people with chronic infection, aging, or immunosuppression.^2,31,32 Chronic CMV reactivation is a potent driver of immune activation through persistent innate immune activation and chronic inflammation.^33,34 Chronic EBV reactivation also promotes sustained immune activation and creates conditions that favor cancer development, particularly when immune surveillance is impaired.³⁵ On the other hand, unlike the herpesviruses, GBV-C has been associated with reduced HIV disease progression and lower inflammation.^7,36

In this large study of HHV reactivation and GBV-C coinfection among MSM with and without HIV-1 spanning 25 years, we observed that plasma samples with detectable HIV-1 RNA had considerably higher odds of CMV, EBV, and HHV-8 viremia compared with samples from men with virally suppressed HIV-1, which in turn had higher odds relative to samples from MWOH. In this study, higher HIV-1 viral load and lower CD4⁺ T cell count, which are markers of HIV-1 disease progression and immune status, were independently associated with presence of viremia for CMV, EBV, and HHV-8. Furthermore, the quantity of CMV was higher in MWH compared with MWOH. Consistent with previous reports,^37–39 our results support that CMV, EBV, or HHV-8 reactivation is associated with high HIV-1 load and disease progression.

Chronic immune activation with high levels of inflammatory cytokines often leads to CD4⁺ T cell activation. Activated CD4⁺ T cells are more susceptible to HIV-1 infection and replication. We found that multiple markers of inflammation and immune activation were present in higher concentrations during CMV and EBV reactivation, and most of these markers were common to both viruses. Both HIV-1 positive and HIV-1 negative samples showed higher concentrations of inflammatory markers with EBV reactivation. However, during CMV reactivation, only HIV-1 positive samples showed evidence of increased inflammation and immune activation. This finding reinforces the results of a previous, smaller study that found associations between CMV antibodies and inflammatory biomarkers among women with HIV-1, but not among women without HIV-1.⁴⁰ Our results indicate that CMV reactivation plays a more synergistic role in HIV-1 pathogenesis.

Our results showed that CRP concentrations, along with those of other markers, were higher among HIV-1 positive samples with CMV viremia. CRP is an inflammatory marker produced by the liver in response to inflammation. Elevated CRP levels have been associated with subclinical CMV reactivation, chronic inflammation and associated morbidities.^41,42 Our results suggest that CMV reactivation is associated with a high level of CRP production and a heightened inflammatory state, which provides a more conducive environment for HIV-1 replication.

To gain insight as to whether men with reactivated CMV and EBV might experience chronic increases in systemic inflammation with age, despite effective HIV-1 suppression, we modeled trajectories of selected inflammatory biomarkers with age among men with consistently undetectable HIV-1 RNA. However, it did not appear that trajectories were meaningfully different among those who reactivated CMV or EBV. This suggests that, barring complications arising from reactivation events, inflammatory spikes are limited to the duration of viremia.

Our results are consistent with prior reports^39,43–46 showing increased risks of CMV, EBV, and/or HHV-8 reactivation in persons with detectable HIV-1 loads. However, viral coinfection may not always be synergistic in HIV-1 disease development. Some reports have shown evidence that HHV-6^47,48 or GBV-C^7,49 coinfection reduces HIV-1 replication in CD4⁺ T cells and slows HIV-1 disease progression. In our results, MWH had lower prevalence of both HHV-6 and GBV-C relative to MWOH. However, there were no detectable differences in prevalence of HHV-6 or GBV-C by HIV-1 viral load or by CD4⁺ T cell count in MWH.

The primary strengths of this study are its size and longitudinal design. Over 11,000 samples provided by 1,882 men were assayed for these five viruses. Participants provided repeated samples over time and across a range of HIV-1-related disease states, allowing for a detailed assessment of relationships among HIV-1, viral coinfections, and markers of inflammatory processes. The inclusion in the MACS of MSM without HIV-1, drawn from the same source population as MWH, allowed us to consider associations with HIV-1 infection while reducing the possibility of confounding, though unmeasured confounding may remain. While most prior studies examined CMV, EBV, or HHV-8 independently, our multi-virus approach in this study offers a more comprehensive understanding of how the interacting burden of CMV, EBV, HHV-8, HHV-6, and GBV-C coinfection relates to inflammation and HIV disease progression in people with HIV—providing insights that single-virus studies cannot capture.

This study has limitations. HHV have high seroprevalence in the overall population (e.g., seroprevalence of HHV-6 approaches 100% by age 3 in some populations),^50,51 and we assumed that detectable viremia represented reactivation rather than primary infection. To bolster this assumption, we used previous results from HHV-8 antibody testing and performed CMV, EBV, and HHV-6 antibody testing on baseline study samples; additionally, we performed antibody testing on all follow-up samples for men who initially tested negative for HHV antibodies. We identified a small number of men whom we suspect acquired primary HHV infection while under follow-up (Supplementary Table S2b). However, there were too few of these men to make inferences about their characteristics, the primary infections themselves, or their sequelae. In addition, low levels of HHV DNA in blood plasma may come from the release of DNA from latently infected cells and not represent reactivation.⁵² We classified the 281 plasma samples (2.7%) with missing HIV-1 load values as detectable or undetectable according to ART use. Although we consider this assumption reasonable, it may have introduced some classification error; however, sensitivity analyses omitting these samples produced extremely similar results with no changed inferences. Study participants provided samples across varying time spans, not always at regular intervals, and with at least 6 months between each visit, precluding detailed inference about the duration of reactivation events and their frequency.

Conclusions

In this comprehensive longitudinal study, we observed a high prevalence of herpesvirus reactivation in the blood of MSM, with reactivation strongly related to HIV-1 viremia. We also found that CMV and EBV reactivations were associated with higher concentrations of markers of inflammation and immune activation, and that the associations with CMV reactivation only applied to HIV-1 viremic samples. In view of the important role of CMV and EBV reactivation in HIV-1 pathogenesis, our results contribute to clinical management and treatment of PWH with and without ART.

Ethics Approval and Consent to Participate

The study was conducted with institutional review board approvals from University of Pittsburgh, Northwestern University, University of California, Los Angeles, and Johns Hopkins University.

Authors’ Contributions

N.I.W.: Data curation and analysis, writing, original draft preparation, editing; Y.C.: Cohort designations, methodology, experimentation, reviewing, and editing; A.B.: Technical dataset and lab experimentation; N.S.-E.: Lab experimentation; C.S.: Lab experimentation; A.D’S.: Reviewing and editing; M.E.: Reviewing and editing; J.H.B.: Reviewing and editing; E.-Y.K.: Reviewing and editing; C.R.R.: Conceptualization, methodology, writing, reviewing, and editing. All authors read and approved the final article.

Footnotes

Acknowledgments

The authors gratefully acknowledge the contributions of the study participants and dedication of the staff at the MACS sites. The authors would like to thank Lori Caruso, Ming Ding, Kathy Kulka, Peter Shoucair, and Susan McQuiston for laboratory technical support.

Author Disclosure Statement

The authors declare that they have no competing interests. The authors declare that they have no relevant or material financial interests that relate to the research described in this paper.

Funding Information

Data in this manuscript were collected by the Multicenter AIDS Cohort Study (MACS), now the MACS/WIHS Combined Cohort Study (MWCCS). The contents of this publication are solely the responsibility of the authors and do not represent the official views of the National Institutes of Health (NIH). MWCCS (Principal Investigators): Baltimore CRS (Todd Brown and Joseph Margolick), U01-HL146201; Data Analysis and Coordination Center (Gypsyamber D’Souza, Stephen Gange and Elizabeth Topper), U01-HL146193; Chicago-Northwestern CRS (Steven Wolinsky, Frank Palella, and Valentina Stosor), U01-HL146240; Los Angeles CRS (Roger Detels and Matthew Mimiaga), U01-HL146333; Pittsburgh CRS (Jeremy Martinson and Charles Rinaldo), U01-HL146208. The MWCCS is funded primarily by the National Heart, Lung, and Blood Institute (NHLBI), with additional co-funding from the Eunice Kennedy Shriver National Institute of Child Health & Human Development (NICHD), National Institute on Aging (NIA), National Institute of Dental & Craniofacial Research (NIDCR), National Institute of Allergy and Infectious Diseases (NIAID), National Institute of Neurological Disorders and Stroke (NINDS), National Institute of Mental Health (NIMH), National Institute on Drug Abuse (NIDA), National Institute of Nursing Research (NINR), National Cancer Institute (NCI), National Institute on Alcohol Abuse and Alcoholism (NIAAA), National Institute on Deafness and Other Communication Disorders (NIDCD), National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), National Institute on Minority Health and Health Disparities (NIMHD), and in coordination and alignment with the research priorities of the National Institutes of Health, Office of AIDS Research (OAR). MWCCS data collection is also supported by UL1-TR000004 (UCSF CTSA), UL1-TR003098 (JHU ICTR), UL1-TR001881 (UCLA CTSI), P30-AI-050409 (Atlanta CFAR), P30-AI-073961 (Miami CFAR), P30-AI-050410 (UNC CFAR), P30-AI-027767 (UAB CFAR), P30-AI-124414 (ERC-CFAR), P30-MH-116867 (Miami CHARM), UL1-TR001409 (DC CTSA), KL2-TR001432 (DC CTSA), and TL1-TR001431 (DC CTSA).

Supplemental Material

References

Cohen

. Herpesvirus latency. J Clin Invest, 2020; 130(7):3361–3369.

Gianella

, Massanella

, Wertheim

, Smith

. The sordid affair between human herpesvirus and HIV. J Infect Dis, 2015; 212(6):845–852.

Ong

DSY

, Bonten

MJM

, Spitoni

, et al.; Molecular Diagnosis and Risk Stratification of Sepsis Consortium. Epidemiology of multiple herpes viremia in previously immunocompetent patients with septic shock. Clin Infect Dis, 2017; 64(9):1204–1210.

Soderberg-Naucler

, Fish

, Nelson

. Reactivation of latent human cytomegalovirus by allogeneic stimulation of blood cells from healthy donors. Cell, 1997; 91(1):119–126.

Forte

, Zhang

, Thorp

, Hummel

. Cytomegalovirus latency and reactivation: An intricate interplay with the host immune response. Front Cell Infect Microbiol, 2020; 10:130.

Murata

, Sugimoto

, Inagaki

, et al. Molecular basis of Epstein-Barr virus latency establishment and lytic reactivation. Viruses, 2021; 13(12):2344.

Williams

, Klinzman

, Yamashita

, et al. Persistent GB virus C infection and survival in HIV-infected men. N Engl J Med, 2004; 350(10):981–990.

Nunnari

, Nigro

, Palermo

, et al. Slower progression of HIV-1 infection in persons with GB virus C co-infection correlates with an intact T-helper 1 cytokine profile. Ann Intern Med, 2003; 139(1):26–30.

Xiang

, Wünschmann

, Diekema

, et al. Effect of coinfection with GB virus C on survival among patients with HIV infection. N Engl J Med, 2001; 345(10):707–714.

10.

Bjorkman

, Flamholc

, Naucler

, Molnegren

, Wallmark

, Widell

. GB virus C during the natural course of HIV-1 infection: Viremia at diagnosis does not predict mortality. AIDS, 2004; 18(6):877–886.

11.

Ryt-Hansen

, Katzenstein

, Gerstoft

, Eugen-Olsen

. No influence of GB virus C on disease progression in a Danish cohort of HIV-infected men. AIDS Res Hum Retroviruses, 2006; 22(6):496–498.

12.

Stern

, Miller

, Li

, Saxena

. Virome and bacteriome: Two sides of the same coin. Curr Opin Virol, 2019; 37:37–43.

13.

Suarez-Zdunek

, Knudsen

, Fuchs

, et al. Cytomegalovirus antibodies and coronary artery disease in people with HIV: A cohort study. Viruses, 2025; 17(2):231.

14.

Kostare

, Kostares

, et al. Prevalence of HHV-6 detection among people living with HIV: A systematic review and meta-analysis. Viruses, 2025; 17(4):531.

15.

Jett

, Tariq

, Yee

. Laboratory characterization of co-infections in individuals infected with HHV-8. Viruses, 2025; 17(4):460.

16.

Stinn

, Kuntz

, Varon

, et al. Subclinical genital herpes shedding in HIV/Herpes simplex Virus 2-Coinfected women during antiretroviral therapy is associated with an increase in HIV tissue reservoirs and potentially promotes HIV evolution. J Virol, 2020; 95(1):e01606–e01620.

17.

Stowe

, Kozlova

, Yetman

, Walling

, Goodwin

, Glaser

. Chronic herpesvirus reactivation occurs in aging. Exp Gerontol, 2007; 42(6):563–570.

18.

Kaslow

, Ostrow

, Detels

, Phair

, Polk

, Rinaldo

Jr . The multicenter AIDS cohort study: Rationale, organization, and selected characteristics of the participants. Am J Epidemiol, 1987; 126(2):310–318.

19.

Detels

, Jacobson

, Margolick

, et al. The multicenter AIDS cohort study, 1983 to. Public Health, 2012; 126(3):196–198.

20.

Wada

, Jacobson

, Margolick

, et al. The effect of HAART-induced HIV suppression on circulating markers of inflammation and immune activation. AIDS, 2015; 29(4):463–471.

21.

Bekker

L-G

, Beyrer

, Mgodi

, et al. HIV infection. Nat Rev Dis Primers, 2023; 9(1):42.

22.

Tseng

, Seet

, Phillips

. The evolution of three decades of antiretroviral therapy: Challenges, triumphs and the promise of the future. Br J Clin Pharmacol, 2015; 79(2):182–194.

23.

Hultin

, Menendez

, Hultin

, et al. Assessing immunophenotyping performance: Proficiency-validation for adopting improved flow cytometry methods. Cytometry B Clin Cytom, 2007; 72(4):249–255.

24.

Agudelo-Hernandez

, Chen

, Bullotta

, et al. Subclinical herpesvirus shedding among HIV-1-infected men on antiretroviral therapy. AIDS, 2017; 31(15):2085–2094.

25.

Ding

, Bullotta

, Caruso

, Gupta

, Rinaldo

, Chen

. An optimized sensitive method for quantitation of DNA/RNA viruses in heparinized and cryopreserved plasma. J Virol Methods, 2011; 176(1–2):1–8.

26.

Balasuriya

, Leutenegger

, Topol

, McCollum

, Timoney

, MacLachlan

. Detection of equine arteritis virus by real-time TaqMan reverse transcription-PCR assay. J Virol Methods, 2002; 101(1–2):21–28.

27.

Niesters

. Clinical virology in real time. J Clin Virol, 2002; 25(Suppl 3):S3–S12.

28.

Rydze

, Bhattarai

, Stapleton

. GB virus C infection is associated with a reduced rate of reactivation of latent HIV and protection against activation-induced T-cell death. Antivir Ther, 2012; 17(7):1271–1279.

29.

Labo

, Miley

, Kesler

, Jacobson

, Whitby

. Current epidemiology of KSHV infection in American men who have sex with men. Infect Agents Cancer, 2010; 5(S1):A13.

30.

Wada

, Bream

, Martínez-Maza

, et al. Inflammatory biomarkers and mortality risk among HIV-Suppressed men: A multisite prospective cohort study. Clin Infect Dis, 2016; 63(7):984–990.

31.

Chaillon

, Nakazawa

, Rawlings

, et al. Subclinical Cytomegalovirus and Epstein-Barr Virus Shedding Is Associated with Increasing HIV DNA Molecular Diversity in Peripheral Blood during Suppressive Antiretroviral Therapy. J Virol, 2020; 94(19):e00927–e00920; doi: 10.1128/jvi.00927-00920

32.

Basso

, Andreis

, Scaggiante

, et al. Cytomegalovirus, Epstein-Barr virus and human herpesvirus 8 salivary shedding in HIV positive men who have sex with men with controlled and uncontrolled plasma HIV viremia: A 24-month longitudinal study. BMC Infect Dis, 2018; 18(1):683.

33.

Moseley

, Klenerman

, Kadambari

. Indirect effects of cytomegalovirus infection: Implications for vaccine development. Rev Med Virol, 2023; 33(1):e2405.

34.

Fowler

, Mucha

, Neumann

, et al. A systematic literature review of the global seroprevalence of cytomegalovirus: Possible implications for treatment, screening, and vaccine development. BMC Public Health, 2022; 22(1):1659.

35.

Ebrahimi

, Rasizadeh

, Sharaflou

, et al. Coinfection of EBV with other pathogens: A narrative review. Front Virol, 2024; 4:1482329.

36.

Tillmann

, Manns

. GB virus-C infection in patients infected with the human immunodeficiency virus. Antiviral Res, 2001; 52(2):83–90.

37.

Deayton

, Prof Sabin

, Johnson

, Emery

, Wilson

, Griffiths

. Importance of cytomegalovirus Viraemia in risk of disease progression and death in HIV-infected patients receiving highly active antiretroviral therapy. Lancet, 2004; 363(9427):2116–2121.

38.

Whitby

, Howard

, Tenant-Flowers

, et al. Detection of Kaposi sarcoma associated herpesvirus in peripheral blood of HIV-infected individuals and progression to Kaposi’s sarcoma. Lancet, 1995; 346(8978):799–802.

39.

da Silva

, de Oliveira

. HIV, EBV and KSHV: Viral cooperation in the pathogenesis of human malignancies. Cancer Lett, 2011; 305(2):175–185.

40.

Lurain

, Hanson

, Hotton

, Weber

, Cohen

, Landay

. The association of human cytomegalovirus with biomarkers of inflammation and immune activation in HIV-1-Infected women. AIDS Res Hum Retroviruses, 2016; 32(2):134–143.

41.

Ridker

, Hennekens

, Buring

, Rifai

. C-reactive protein and other markers of inflammation in the prediction of cardiovascular disease in women. N Engl J Med, 2000; 342(12):836–843.

42.

Simanek

, Dowd

, Pawelec

, Melzer

, Dutta

, Aiello

. Seropositivity to cytomegalovirus, inflammation, all-cause and cardiovascular disease-related mortality in the United States. PLoS One, 2011; 6(2):e16103.

43.

Webster

, Lee

, Cook

, et al. Cytomegalovirus infection and progression towards AIDS in haemophiliacs with human immunodeficiency virus infection. Lancet, 1989; 2(8654):63–66.

44.

Webster

. Cytomegalovirus as a possible cofactor in HIV disease progression. J Acquir Immune Defic Syndr (1988), 1991; 4(Suppl 1):S47–S52.

45.

Griffiths

. CMV as a cofactor enhancing progression of AIDS. J Clin Virol, 2006; 35(4):489–492.

46.

Emery

, Griffiths

. Prediction of cytomegalovirus load and resistance patterns after antiviral chemotherapy. Proc Natl Acad Sci U S A, 2000; 97(14):8039–8044.

47.

Grivel

, Ito

, Fagà

, et al. Suppression of CCR5-but not CXCR4-tropic HIV-1 in lymphoid tissue by human herpesvirus 6. Nat Med, 2001; 7(11):1232–1235.

48.

Kainth

, Fisher

, Fernandez

, et al. Understanding the association between chromosomally integrated human herpesvirus 6 and HIV disease: A cross-sectional study. F1000Res, 2013; 2:269.

49.

Bhattarai

, Stapleton

. GB virus C: The good boy virus? Trends Microbiol, 2012; 20(3):124–130.

50.

Braun

, Dominguez

, Pellett

. Human herpesvirus 6. Clin Microbiol Rev, 1997; 10(3):521–567.

51.

Okuno

, Takahashi

, Balachandra

, et al. Seroepidemiology of human herpesvirus 6 infection in normal children and adults. J Clin Microbiol, 1989; 27(4):651–653.

52.

Agut

, Bonnafous

, Gautheret-Dejean

. Laboratory and clinical aspects of human herpesvirus 6 infections. Clin Microbiol Rev, 2015; 28(2):313–335.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.04 MB

0.03 MB