Highly active antiretroviral therapy (HAART) or combination antiretroviral therapy (cART) has significantly increased life expectancy of people living with HIV/AIDS (PLWHA) in the past decades . This has resulted in a shift in patterns of morbidities and mortalities in this population. A remarkable note is that both AIDS-defining cancers (ADCs) and non-AIDS-defining cancers (NADCs) are more often seen among PLWHA than among the general population [Reference Pantanowitz, Schlecht and Dezube2–Reference Engels4]. PLWHA are at long-term status of immunodeficiency, immune activation and inflammation caused by HIV infection [Reference Younas5, Reference Ipp6], and have been observed to be at higher risks for both ADCs and NADCs than general population even after cART [Reference Wada7–Reference Silverberg9]. In addition to higher prevalence of infection with cancer-related viruses such as hepatitis C virus (HCV), human papilloma virus (HPV), Kaposi sarcoma associated herpes virus (KSHV) among PLWHA, higher prevalence of certain behavioral and psychological cancer risk factors such as tobacco use, heavy drinking and depression, etc. also plays important roles in carcinogenesis in this population. Thus, surveillance on cancer incidence and mortality as well as risk factors in this population are important for learning the disease burden and targeting people at high risk for cancer. Moreover, the associations between risk factors and carcinogenesis might be different in this immune compromised population, appealing for more detailed exploration.
The Epstein–Barr virus (EBV) is from the herpes virus family and persistently infects more than 90% of human adults. EBV infects B cells and epithelial cells with different mechanisms and replication patterns, and shows life-long latency after the initial lytic period [Reference Amon and Farrell10]. Although the infection is ubiquitous and usually asymptomatic, EBV is constantly associated with certain types of human cancers including Hodgkin's lymphoma (HL) [Reference Punnett, Tsang and Hodgson11], non-Hodgkin's lymphoma (NHL) [Reference Pietersma, Piriou and van Baarle12, Reference Molyneux13] and nasopharyngeal cancer (NPC) [Reference Busson14, Reference Chua15], among which NHL is one of the three ADCs. It is believed that genetic characteristics from both the viral and host sides, diet, behavioral, and environmental risk factors interplay in EBV-associated cancer development. In addition to EBV infection, risk factors for NPC include HPV infection, male, several dietary and behavioral factors such as salted fish intake and smoking, and family history of NPC [Reference Guo16, Reference Xue17]; potential risk factors for HL include age, male, weakened immune system, and family history of the disease ; and risk factors for NHL vary by subtypes, including male, HIV infection, oncogene activation (c-MYC), malaria endemicity, family history, cigarette smoking, alcohol consumption, autoimmune diseases, and occupational factors [Reference Molyneux13, Reference Chihara19].
EBV DNA can be detected in both tumor tissues and blood samples from patients with EBV-related cancers, and circulating viral DNA level was proved to be an important biomarker in risk stratification, disease progression, and prognosis for NPC [Reference Chua15, Reference Han20, Reference Du21] and lymphoma [Reference Lei22, Reference Park23]. It has been reported that the EBV DNA were detected more often among HIV infected individuals than negative ones, suggesting a more active viral replication [Reference van Baarle24]. However, little has been explored for the association between non-viral risk factors and circulating EBV DNA in understanding risk and progression of EBV-related cancers especially in PLWHA.
Approximately 575 000 people were living with HIV/AIDS in China at the end of 2015. Higher prevalence and mortality of cancers have already been observed in this population [Reference Yang25]. Additionally, homosexual transmission of HIV between men has been markedly increasing in recent years, reporting a HIV prevalence of 8% in 2015 . High prevalence of HIV infection and risky behaviors made it an urgent need to have close observation on cancer development in this population.
We hypothesize that HIV infection and other potential risk factors of EBV-related cancers may be associated with EBV DNA replication. This study was thus performed to compare whole blood EBV DNA positivity between HIV-positive and HIV-negative men who have sex with men (MSM), as well as to identify potential correlates with EBV DNA positivity in Shanghai, China.
All HIV-positive MSM and 1:1 age-matched (±5 years) HIV-negative MSM were recruited from an HIV counseling and testing clinic in Shanghai from November 2014 to November 2015 and were administered with a questionnaire interview. The study was approved by the Institutional Review Board of Fudan University and written informed consent was obtained from each participant.
Subjects who met the following inclusion criteria were enrolled: (1) male; (2) 18 years old or above; (3) can give oral or written informed consent in Mandarin; and (4) had oral or anal sex with men within the past 12 months.
Participants were administered with a face-to-face questionnaire interview by trained and experienced health professionals to ascertain socio-demographic characteristics including age, residency, ethnic group, education level, marital status, and monthly income; and behavioral characteristics including lifetime homosexual behaviors, number of male sexual partners, tobacco smoking, alcohol drinking, dietary habits especially on unhealthy food intake, and history of illicit drug use. HIV-related clinical characteristics such as CD4+ T cell counts, diagnosis of AIDS and status of receiving ART were obtained from the Chinese National Information System for AIDS Prevention and Control (CNISAPC) given that all PLWHA in China have registered with and been routinely followed-up by CNISAPC.
Participants were asked to report their frequency of pickled, smoked, or salty food intake, and frequency of having fresh vegetables or fruits in the past 6 months. Frequencies were reported as ‘never’, ‘occasionally’, or ‘frequently or daily’.
For the measurement of tobacco smoking, alcohol drinking and illicit drug use, the definitions were as following. Smoked in the past 6 months was defined as had smoked at least one cigarette every day during that period. Drank in the past 6 months was defined as had drunk alcohol at least once per week during the past 6 months. Types of alcohol consumed include beer, yellow wine, rice wine, Chinese white wine (a kind of distilled spirit), red wine, white wine, whisky, brandy, etc. Illicit drugs use in the past 6 months was defined as used methamphetamine, ecstasy, heroin, marijuana, opium, cocaine, sedatives, ketamine or poppers (amyl nitrites) during the past 6 months.
Participants were asked about sexual behaviors and the frequency of condom use. And the total number of sexual partners in life time was also collected.
EBV DNA extraction, amplification, and sequencing
For each participant, 5 ml venous blood withdrawn using an EDTA anticoagulation sterilized tube were further aliquoted in both whole blood and plasma, respectively. EBV-DNA was extracted from 200 µl whole blood in a nucleic acid-free room by High Pure Viral Nucleic Acid Kit (Roche, Germany) according to the manufacturer's instruction. The DNA was eluted into 50 µl elution buffer provided by the kit and stored at −20 °C until testing by nested polymerase chain reaction (PCR).
Nested PCR was performed to amplify the EBNA-1, EBNA-2, and LMP-1 genes of EBV. PCR reaction system was prepared in another nuclease-free room. The reaction mixture contained a final concentration of 1×PCR buffer, 200 µM of each dNTP, 0·4 µM of each primer, and 1·25 U of Taq polymerase. Outer and inner primers used for EBNA-2, EBNA-1, and LMP-1 genes in this study are shown in Table 1. About 1 µl of DNA (25–30 ng) was added into the mixture with aerosol-resistant pipette tips to avoid cross-pollution between samples in the nucleic acid extraction room. PCR amplification was performed in a different room to avoid pollution. The integrity of the extracted DNA and the exclusion of PCR inhibitors in samples were confirmed by amplifying the human gene. The first amplification round included an initial denaturation step at 95 °C for 5 min, followed by 30 cycles of denaturation at 95 °C for 1 min, primers annealing at 50 °C (EBNA-2)/47 °C (EBNA-1)/52 °C(LMP-1) for 1 min, and an extension step at 72 °C for 1 min, followed by the final extension step at 72 °C for 10 min. For the second round, 1 µl of the first round PCR product was used as template and amplified under the same conditions. The second round of amplification products were tested in 2% agarose gel. The positive PCR products were purified and sequenced by the Beijing Genomics Institute (BGI), Beijing, China. The participant with any one of the above three target genes detected was defined as positive for EBV DNA.
Statistical analyses were performed using R (https://www.r-project.org/). The categorical variables were compared using χ 2 test or Fisher's Exact test when appropriate. Unconditional logistic regression was used to estimate the odds ratios and their 95% confidence intervals. The selection of variables into multiple logistic regression models was based on both its significance in the univariate regression and prior knowledge (for socio-demographic characteristics and potential known risk factors). A two-sided P-value <0·05 is considered as statistically significant.
A total of 1008 participants including 504 HIV-positive and 504 HIV-negative MSM were included in the study. Among them, 81·7% aged at 21–40 years (range 18–69 years), 62·4% were non-local Shanghai residents, 96·0% were ethnic Han, 66·8% received college or higher education, 76·5% were never married, and 50·9% reported to have monthly income more than 5000 RMB yuan (~US$735). The two groups were comparable in distributions of residency, ethnicity, and marital status, but were significantly different in age, education level, and monthly income (Table 2).
Lifetime homosexual behaviors
As shown in Table 2, 95·4% of the participants had homosexual behavior with a main male partner, 17·2% with a male sex worker, and 79·1% with a casual male partner in their lifetime. The majority (95·8%) of the participants reported having had two or more sexual partners and 41·6% having had ⩾10 partners so far. Compared with HIV-negative controls, HIV-positive MSM reported significantly higher proportion of homosexual activities including unprotected casual sex and having ⩾5 partners.
The majority (87·5%) of the participants reported having had pickled, smoked, or salty food at least once in the past 6 months, and 7·5% had such food frequently or daily. Also 27·4%, 37·9%, and 30·5% of participants reported smoking tobacco, drinking alcohol and illicit drug use in the past 6 months. Forty percent had depressive symptoms when measured with the CES-D (Center for Epidemiologic Studies Depression Scale). Significantly higher proportions of HIV-positive MSM reported fresh vegetables or fruits intake, smoking, drugs use, and depression than HIV-negative group, but a lower proportion consumed alcohol (Table 2).
HIV infection-related characteristics
Among the 504 HIV-positive MSM, 142 (28·2%) had been diagnosed with AIDS, 300 (59·5%) were receiving ART, and 90 (17·9%) showed CD4+ T cell counts <200 cells/μl at the time of interview.
Prevalence and correlates of whole blood EBV DNA
The whole blood EBV DNA positivity was 56·0% (95% CI 51·6–60·2%) for HIV-positive MSM and 26·0% (95% CI 22·4–30·0%) for HIV-negatives (χ 2 = 93·53, P < 0·001).
In the univariate logistic regression analyses for all participants, positivity of whole blood EBV DNA was inversely associated with levels of education (crude odds ratio (cOR) = 0·55, 95% CI 0·38–0·82), but was positively associated with aged above 40 years old (cOR = 1·72, 95% CI 1·21–2·45), ever married (cOR = 1·43, 95% CI 1·07–1·92), ever had commercial sex with a male sex worker without consistent condom use in life time (cOR = 1·59, 95% CI 1·09–2·31), ever had sex with a casual male partner without consistent condom use in lifetime (cOR = 1·55, 95% CI 1·13–2·14), had five or more male partners in life time (cOR = 1·53, 95% CI 1·12–2·10), frequent or daily intake of pickled, smoked, or salty food in the past 6 months (cOR = 1·98, 95% CI 1·23–3·17), used illicit drugs in the past 6 months (cOR = 1·45, 95% CI 1·10–1·90), had depression symptoms (cOR = 1·31, 95% CI 1·01–1·69) and HIV infection (cOR = 3·62, 95% CI 2·77–4·72) (Table 3). In the multiple logistic regression analysis with adjustment for potential confounding variables, positivity of whole blood EBV DNA was positively associated with HIV infection (adjusted odds ratio (aOR) = 3·43, 95% CI 2·58–4·57) and with frequent intake of pickled, smoked, or salty food (aOR = 1·71, 95% CI 1·02–2·86) (Table 3).
a Adjusted for all the other variables in the table.
The associations were further examined stratified by status of HIV infection and were shown in Table 4. For HIV-infected MSM, 282 were tested positive for whole blood EBV DNA. Positivity of whole blood EBV DNA was positively associated with CD4+ T cell counts <200 cells/μl (aOR = 1·79, 95% CI 1·05–3·05) and with frequent intake of pickled, smoked, or salty food (aOR = 3·14, 95% CI 1·39–7·08). For HIV-negative MSM, 131 were tested positive for whole blood EBV DNA and the positivity was only significantly associated with monthly income (3000–4999 vs. <1000 RMB yuan: aOR = 3·75, 95% CI 1·21–11·65; ⩾5000 vs. <1000 RMB yuan: aOR = 3·61, 95% CI 1·21–10·71) in the multiple logistic regression model.
a Adjusted for all the other variables in the table.
In this cross-sectional study, we reported a higher prevalence of whole blood EBV DNA among HIV-positive participants and its positive association with dietary factors and immune-suppressed status in a MSM population from Shanghai, China. These results suggested an active EBV replication status among PLWHA, and appealed for further explorations in the risk factors and their interaction in EBV-related carcinogenesis.
The prevalence of whole blood EBV DNA was found to be 56·0% (95% CI 51·6–60·2%) for HIV-positive MSM and 26·0% (95% CI 22·4–30·0%) for HIV-negative MSM in this study. Among limited literature in Chinese PLWHA, Wu et al. reported an overall prevalence of peripheral EBV DNA as 32·3% in 257 HIV/AIDS patients from a hospital in Shenzhen [Reference Wu31]. Wang et al. reported higher prevalence among 73 HIV-infected participants when compared with 80 healthy controls in Henan (35·6% vs. 5·0%) [Reference Wang, Chen and He32]. The prevalence for HIV-infected people found in the current study was higher than the above mentioned studies, probably because it targeted HIV-positive MSM population specifically. Evidence from a study performed in Amsterdam showed a prevalence of 67% for EBV DNA from peripheral blood among HIV-positive MSM, higher than the prevalence of 39% among HIV-negative MSM and the prevalence of 6% among HIV-negative heterosexual men [Reference van Baarle24]. Our results added to the evidence that HIV-positive homosexual men were at higher risk of having positive whole blood EBV DNA.
The higher prevalence of EBV DNA among PLWHA indicated that people with suppressed immune system were at higher risk for active EBV viral replication, which was also suggested by our finding that having CD4+ T cell counts <200 cells/ul were positively associated with detectable EBV DNA. Wu's study also found higher EBV DNA prevalence among inpatients who were at advanced stage of AIDS than outpatients (44·0% vs. 25·9%, χ 2 = 8·7605, P < 0·005) [Reference Wu31]. Chronic HIV infection causes antigen simulation of B cells, breaks the equilibrium that newly infected and differentiating B cells were controlled by cytotoxic T cell responses, and thus leads to uncontrolled lymphoproliferation and cancer development [Reference Pietersma, Piriou and van Baarle12]. It has been observed that the standardized incidence rates for most NADCs were greater than doubled in HIV population, and lower CD4+ T cells and taking HAART were found to be two major risk factors [Reference Rubinstein, Aboulafia and Zloza33]. Moreover, it has been believed that the EBV DNA load and HIV viremia are correlated [Reference Petrara34] and could serve as the diagnostic and prognostic marker for lymphoma in HIV-infected patients [Reference Navarro35–Reference Piriou38]. Based on these evidence and our finding, we suggest closer surveillance of EBV-related cancer incidence, known risk factors, as well as biomarkers including EBV DNA especially among PLWHA with lower CD4 counts or with HIV viremia.
One major finding of the current study was that participants who frequently consume pickled, smoked, or salty food were at higher risk of having positive whole blood EBV DNA. Among known risk factors for NPC, salted fish intake was a well-established one and was firstly reported in Chinese population [Reference Ho, Huang and Fong39, Reference Ning40]. Chinese-style salted fish was defined as Group I carcinogen by IARC (International Agency for Research on Cancer) since high levels of NDMA (N-nitrosodimethylamine) was generated during making the fish. In this study, we were the first to investigate participants’ habit of pickled, smoked, or salty food intake, which was a broad range of food processed using salt or other sources, and found a positive association with circulating EBV DNA in MSM. In our observation, the association was only significant among HIV-positive MSM in stratified analyses. The possible explanations for the heterogeneity of the association include that HIV infection could be a stronger risk factor for active EBV replication, making those under immunosuppressed condition more vulnerable to other potential cancer risk factors to play a role. This is consistent with our observation that lower CD4 T cell counts also showed to be a risk factor. Once confirmed, the association between pickled, smoked or salty food and active EBV replication would suggest a population at higher risk of EBV-related cancers for intervention and cancer prevention.
There are several limitations in the current study. First, by nature of a cross-sectional design, we may only describe the distribution of risk factors and associations identified in this study can be further investigated using other study designs. However, the associations found between positive whole blood EBV DNA and HIV infection and frequent pickled, smoked, or salty food intake still calls attention not only because it was the first time examining the association in this population, but also for the strong associations discovered. Second, residual confounding may be present and partly explain the association we found between EBV DNA and cancer risk factors. For example, HPV infection is a risk factor for NPC and has not been measured in this study. We collected information on sexual behaviors as a proxy, since HPV infection is associated with active and unsafe sex. A more comprehensive collection of information on exposures for known risk factors such as viral co-infection is needed in future research. Third, most of the exposure measurements from this study were based on self-reported data. Thus the information bias may exist especially for the validity of answers to sensitive questions such as on sexual behaviors and illicit drug use. Also, the measurements on tobacco and alcohol use, dietary patterns, and quantities of food can be improved for better estimates of the associations. However, the validity of the major finding on HIV infection and unhealthy food intake and their associations with EBV DNA would not be biased seriously since HIV infection was determined by laboratory tests and questions on dietary were not sensitive. Last but not least, although we included all the HIV-infected MSM attending the clinic with matched controls during the study period, the prevalence and the association may not be representative for all MSM in Shanghai or in China.
In conclusion, this study reported a prevalence of 56% for positive whole blood EBV DNA among the HIV-infected MSM, higher than the prevalence of 26% among the HIV-negative MSM in Shanghai, China. It also reported significant association between HIV infection, frequent pickled, smoked, or salty food intake, and lower CD4+ T cell counts with detectable whole blood EBV DNA. These results suggest that HIV-positive MSM in Shanghai, China are at higher risk for EBV-related cancers and further conformational studies are warranted.
This study was supported by Natural Science Foundation of China (grant no. 81361120385), Shanghai Municipal Health and Family Planning Commission (grant numbers GWTD2015S05, 15GWZK0101, 15GWZK0801, and 201440328), and the Chinese National Major Science & Technology Project of Infectious Diseases (grant no. 2012ZX10001007–006).
DECLARATION OF INTEREST
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008.