Is Epstein‐Barr Virus Transmitted Sexually?

Herpesviruses are among the most successful viruses in the human population. More often than not they do not even cause disease, yet they persist for life. Their hallmark is the ability to infect quite different primary and secondary cell types, with different outcomes in each: replication of virus in the primary target tissues and an enduring latent state in the secondary infection sites. Herpes simplex virus type 2 (HSV-2), for example, replicates initially in the vagina and cervix and then takes up silent residence in sacral ganglia, whence it can reactivate and cause renewed viral replication and vaginal lesions. Herpesviruses are efficient biologic vectors, and they are versatile in their routes of transmission. Of the 8 human herpesviruses, 3—HSV-2, cytomegalovirus (CMV), and Kaposi sarcoma-associated herpesvirus (KSHV)—are well established as sexually transmitted agents. Why, then, should it be any surprise to find that Epstein-Barr virus (EBV), which we know best as the causative agent of infectious mononucleosis (IM)—usually in postpubescent young people—is probably sexually transmitted? First, let us be clear; all 4 of the above-mentioned herpesviruses are more often transmitted by nonsexual contact, but why has it taken so long to think of EBV as a sexually transmitted agent at least some of the time? The answers are varied and reflect the history of EBV [1]. It was discovered (in 1964) in a distinctive oral lymphoma in African children, Burkitt lymphoma (BL), and was thereafter thought to be a lymphotropic virus. Even though it was detected not long afterward in a nonlymphoid malignancy, nasopharyngeal carcinoma, the idea that the virus could infect epithelial cells was not substantiated until detection of replicating EBV genomes in oropharyngeal epithelial cells in vivo by in situ nucleic acid cytohybridization in 1977 [2] and was verified later [3]. Indeed, an early successful attempt to infect epithelial cells in vitro used cells growing out from explanted plugs of human cervical tissue, but infection was inefficient, probably because the virus prefers differentiated cells in which to replicate, and epithelial cells de-differentiate when cultured ex vivo. Before then, IM was already branded “the kissing disease” in a less sexually venturesome age. Detection of replicating EBV genomes in cervical cells and washings was to come only in 1986 [4], and later in males [5]. However, unlike HSV and CMV infection, EBV infection was not then and is not now associated with cervicitis, whereas it is notorious for the pharyngitis so characteristic of IM. Finally, seroepidemiological evidence of sexual transmission was lacking—it was sought and found only relatively recently by investigators in this issue of the Journal [6], as well as sporadically by others [7]. Even with this large study, proof remains elusive—epidemiologically because questions about oral-genital sex were not posed to the entering Edinburgh University students in an otherwise well-designed retrospective study, and virologically because EBV types are less well defined and organ-compartmented than HSV-2. The biological evidence—EBV replicates in cervical epithelial cells in vitro and in vivo; is easier to detect in the cervix of sexually active women, especially if they have other sexually transmitted diseases (STDs); and is found in male seminal secretions—together with the epidemiological evidence makes a good case for sexual transmission. However, because EBV is excreted copiously in saliva for weeks leading up to onset of IM and for many months thereafter (as well as in asymptomatic infection), direct salivary exchange is the primary mode of transmission in adults, as it is in infants and children. Yet certainly the frequency of sexual transmission of the virus in acute IM has been too little recognized. Silent but situated as it may be in the cervix of parturient women, does EBV cause perinatal infection? Unlike with HSV and CMV, the evidence is scant. Circulating EBV antibodies are a more substantial barrier to neonatal infection than is the case for the other 2 herpesviruses. In regions in which BL is endemic in Africa, the prevalence of EBV infection by the first year of life is ∼90% but seems mostly to get under way by 4–6 months after birth, as titers of maternal antibodies wane. The situation is very different for CMV, which is sexually transmitted with high frequency in adolescent girls and may cause devastating congenital infection in the first trimester, fortunately rarely; more commonly it causes perinatal infection, which then leads to largely silent excretion of virus in saliva and urine and, hence, nursery transmission [8]. EBV is an oncogenic virus, in fact the first human tumor virus identified. Could it—unlike the nononcogenic HSV-2, which was mistakenly linked to cervical carcinoma by faulty seroepidemiological studies a generation ago—contribute to the genesis of cervical cancer? No such evidence has emerged, despite various and admittedly somewhat desultory efforts through the years (nor has it emerged for cytomegalovirus). That EBV can cause cervicitis or contribute to it or vaginal lesions is plausible but essentially unstudied—or at least not reported, except for anecdotal cases [9]. KSHV, the other oncogenic human herpesvirus, is intriguing for the contrasts with EBV it presents. It was, in fact, suspected to exist because the epidemiological profile of Kaposi sarcoma suggested a sexually transmitted viral etiology. The virus is excreted in saliva but is much more difficult to detect in oropharyngeal secretions than EBV. As predicted, KSHV is detected in male and female genital secretions [10, 11]. In contrast to EBV, these sources of the viruses do lead to malignancy (Kaposi sarcoma), both in the immunocompromised and in regions of indigenous KS in Africa and elsewhere, giving prophylaxis high priority. What is the physician to advise sex partners, one of whom has mononucleosis? Although the strength of the evidence is borderline, perhaps because the amounts of virus excreted from the cervix (data on EBV titers in semen are even more meager) are generally much less than those in saliva, use of condoms may make sense. This is more suggestion than admonition, because transmission may well have occurred already during the incubation period. Excretion of EBV in the saliva may continue for years, but titers fall off subacutely in the weeks after onset, probably to noncontagious levels. As to treatment for IM, it has remained quite elusive, despite potent antiviral drugs that stop EBV replication and can reduce viral levels in saliva, probably because most of the symptoms of IM are the consequence of the explosive cytotoxic T cell responses to EBV-infected B cells so characteristic of this illness, rather than to the virus-infected tissues themselves [12]. Thus, of the 8 human herpesviruses, 4 can be sexually transmitted. These 4 viruses—HSV, CMV, KSHV, and EBV—present a spectrum of sexual transmissibility and pathogenic consequences and differ greatly in both respects. Although EBV is clearly a member of the sodality, despite all its other still-unfolding virological, oncological, and sociological fascinations, it is unlikely to be given much priority in STD clinics anytime soon.