Epstein-Barr Virus and the Eye

Epstein-Barr Virus (EBV), also known as Human HerpesVirus-4 (HHV-4), is a γ-type herpesvirus with a global infection prevalence of over 95%.^1,2 Infection, which occurs nasopharyngeally via either aerosol or direct contact, tends to be asymptomatic in children, but may be associated with the development of infectious mononucleous (IM) in 50% or more of adolescents and young adults. Common symptoms of IM including fever, fatigue, sore throat, adenopathy and hepatosplenomegaly. Whether in children or adults, acute infection is invariably followed by lifelong latency, with B-cells serving as the primary EBV reservoir. More limited latency can be established in epithelial, Natural Killer (NK)- and T-cells.

The occurrence of EBV-induced lymphoid or epithelial malignancies is well-recognized.^3–5 The most common EBV-cancers are nasopharyngeal carcinoma, gastric carcinoma, and the B-cell malignancies, including Hodgkin lymphoma, Burkitt lymphoma, diffuse large B cell lymphoma, and primary central nervous system and vitreoretinal lymphoma.^6–8 Uncommon EBV-associated malignancies include leiomyosarcoma, and NK/T-cell lymphomas and leukemias. Of note, EBV-induced NK/T-cell clonal proliferation can produce the systemic syndrome of Chronic Active EBV (CAEBV) infection, a rare lymphoproliferative disorder reported most often in people of Asian or indigenous Central or South American origin and characterized by multiorgan system infiltration and inflammation,^9–11 including uveitis.^12,13 Cutaneous forms of CAEBV characterized by papulovesicular rash and/or severe mosquito bite allergy, both of which can be associated with severe ulceration, necrosis, and scarring, exist as well.⁹ Acute, fulminant, and often fatal, EBV-induced NK/T-cell malignancies have also been described, including Hemophagocytic LymphoHistiocytosis (HLH). Transient and recurrent EBV reactivations occur commonly in the setting of systemic immunotherapy and Human Immunodeficiency Virus (HIV) infection, both of which are associated with an increased risk of EBV-cancers.¹⁴ EBV-associated lymphoproliferation may also occur following solid-organ or allogeneic hematopoietic stem-cell transplants – so-called Post-Transplantation Lymphoproliferative Disorders (PTLD),¹⁵ which can involve the eye.¹⁶

The diagnosis of acute vs latent EBV infection is generally serologic and depends upon the identification of serum antibodies directed against Viral Capsid Antigen (VCA) alone in the setting of acute infection verses those directed against both VCA and EBV Nuclear Antigen (EBNA) in latent disease.¹⁷ In contrast, quantitative polymerase chain reaction (PCR)-based amplification of EBV DNA and in situ hybridization of EBV RNA are generally reserved for diagnosing recurrent, chronic and/or neoplastic forms of EBV-associated disease.¹⁸ No agents have been approved to treat acute or chronic EBV infection in the United States or Europe, although the nucleoside analogues acyclovir, penciclovir, and ganciclovir and their respective pro-drugs have been shown to have anti-EBV activity in vitro and in selected patient populations.^19,20

A direct role for EBV in either the initiation or progression of autoimmune disease has been suggested, but remains controversial.^21,22 On the one hand, EBV DNA, EBV proteins, and anti-EBV antibodies tend to be elevated in patients with a number of autoimmune disorders when compared to healthy controls. Moreover, the patterns of these EVB proteins and anti-EBV antibodies suggest persistent lytic reactivation in many patients. Such observations, together with the fact that a molecular resemblance exists between several EBV proteins and selected self-antigens, has led some experts to propose that EBV reactivation followed by bystander leukocyte activation with epitope spreading and/or molecular mimicry between EBV and self-proteins might contribute to the pathogenesis of autoimmune disease in some patients. On the other hand, however, it is clear that EBV infection is far more prevalent than autoimmune disease; that at least some portion of those with autoimmune disease are not EBV infected or show no evidence of lytic reactivation; and that both autoimmune disease and the immunotherapies used to treat such conditions might readily result in persistent EBV reactivation with increased anti-EBV antibody production. Such complex virus-host interactions need to be weighted carefully, therefore, when seeking to ascribe EBV causality to autoimmune conditions, including ocular inflammation.

Several studies^23-31 have assessed intraocular levels of EBV DNA and/or anti-EBV antibodies in patients with uveitis, providing evidence for the presence of EBV in up to 10% of eyes with active inflammation. The majority of these “EBV-positive eyes” had other, well-accepted causes for uveitis, however, leading recent authors to question the relevance of such EBV-positive results and to conclude ultimately that routine EBV testing in patients with ocular inflammation is neither useful nor recommended.^27–31 Of note from recent studies, moreover, were the repeated observations that patients with uveitis and high intraocular levels of EBV DNA or anti-EBV antibodies appeared to be at increased risk for HIV infection – including several patients known to have primary vitreoretinal lymphoma. Three Original Articles^32-34 and two letters^35,36 in this issue of Ocular Immunology & Inflammation (OII) report either directly or indirectly on the role of EBV in eyes with inflammation.

Keorochana et al³² studied 64 subjects seen at a uveitis referral center in Bangkok, Thailand, between 2008 and 2018. All patients had anterior uveitis, ocular hypertension, and underwent PCR-based analysis for the presence of varicella zoster virus (VZV), herpes simplex virus (HSV; subtype unspecified), cytomegalovirus (CMV), and EBV DNA. Subjects with a known cause for their uveitis or with preexisting glaucoma/ocular hypertension or with previous topical corticosteroid use, intraocular injection, or ocular trauma or surgery (except cataract surgery) were excluded from the study. Details regarding the sensitivity and specificity of the PCR-based assays used, as well as whether or not subjects were on anti-viral agents at the time of testing were not provided. For those with no evidence of VZV, HSV, CMV or EBV DNA in their aqueous fluid, a diagnosis of either Fuchs Uveitis Syndrome (FUS) or Posner Schlossman Syndrome (PSS) was ascribed to those patients who met accepted syndromic criteria.^37–39 Overall, 53.1% of patients were PCR positive, including 34.4% for CMV, 10.9% for VZV, 6.3% for HSV (type not specified), and 1.6% (one) for EBV. The mean age of the cohort was 48.8 years (SD 14.6 years), with those who were PCR-positive tending to be older (52.7 vs 44.4 years; nominal p = .021) and more likely to have pupillary dilation (26.5% vs 6.5%; nominal p = .036) or sector iris atrophy (17.6% vs 3.3%; nominal p = .083), or to develop glaucoma (44.1% vs 16.7%; nominal p = .18) or corneal decompensation (11.8% vs 0%; nominal p = .052) associated with lower mean corneal endothelial cell counts (CECC; 2193 vs 1879; nominal p = .004). When adjusted for confounding using logistic regression, only older age (p = .026) and the presence of pupillary dilation (p = .050) and lower CECC (p = .054) achieved statistical significance. As has been reported from other Asian centers,^40,41 the cohort was notable for the high rate of CMV-associated anterior uveitis, which tended to have a high maximal intraocular pressure (IOP; mean 41.7 ± 11.7 mmHg) and to be associated coin-shape keratic precipitates. Eyes with VZV-associated anterior uveitis were more likely to have decreased corneal sensation and iris changes.⁴² While follow up varied (mean 40.5 months; SD ±103.4 months), overall 70.3% of eyes experienced recurrences. Among the five eyes diagnosed with FUS, all had cataract, 3 (60%) had stellate keratic precipitates, and 2 (40%) developed glaucoma. Among the 16 eyes diagnosed with PSS, four (25%) had iris changes and two (12.5%) had decreased corneal sensation, all suggestive of PCR-negative herpetic disease.^37–39 The single subject found to be EBV-positive on PCR testing was 61-years-old and had mild anterior uveitis with an IOP of 27 mmHg and no keratic precipitates, iris changes, or cataract. The authors acknowledged the limitations of the study, particularly the potential for referral, recall, and selection biases – including the exclusion of immunocompromised patients. The authors concluded that herpes virus infections accounted for roughly half of all cases of hypertensive anterior uveitis in their clinic. While neither mentioned nor discussed, the authors did not distinguish or identify those subjects with Inflammatory Ocular Hypertension Syndrome (IOHS), i.e., elevated IOP that occurs at the onset of active inflammation and which is suggestive of herpetic disease, as opposed to chronically elevated IOP, which is more typical of FUS.^37–39 The significance of the single EBV-positive case was not commented upon.

By way of direct regional comparison, Khieu et al³³ performed a similar study of 31 HLA-B27-negative subjects with hypertensive anterior uveitis seen between 2016 and 2019 in a referral clinic in Chiang Mai, Thailand, using PCR-based testing for CMV, HSV, VZV, rubella virus, chikungunya virus, and Zika virus. Only patients with a negative uveitis workup and who were suspected of having herpetic disease underwent PCR-based testing. Just under one-third (32%) of subjects were PCR-positive, including 19% CMV, 10% HSV, and 3% VZV. As with the study from Bangkok by Keorochana et al,³² Khieu et al³³ concluded that herpes viruses account for a sizable proportion of cases of hypertensive uveitis and that CMV was an especially common cause in Thailand, which was again in line with other studies from Asia.^40,41 Epstein-Barr virus was not tested for in this study.

Mashima et al³⁴ described an 83-year-old woman seen in a referral clinic in Tokyo, Japan, who developed unilateral necrotizing retinitis in the setting of a 20-year history of methylprednisolone treatment for interstitial pneumonia and chronic pyelonephritis. PCR-based testing revealed high EBV DNA levels in the aqueous humor, but not in the plasma. Testing of the aqueous for HSV, VZV, and CMV DNA were all negative, as were serologic tests for antibodies against Toxoplasma gondii, Treponema pallidum, and HIV. Serial intravitreal methotrexate injections elicited an immediate reduction in EBV DNA copy levels along with clinical improvement. One wonders, therefore, whether the retinitis in this patient was neoplastic. The authors cited a similar case by Imai et al.⁴³ that was ultimately found to have intraocular NK/T-cell lymphoma. While several cases of intraocular NK/T-cell lymphoma have been reported previously,^44–47 such rare EBV-associated NK/T-cell tumors appear to be more likely to involve the orbit or ocular adnexa.^48–50

Chen and Xu³⁵ described a 50-year-old man seen at a referral center in Changsha, China who presented with a unilateral conjunctival injection, intraocular inflammation and inferior exudative retinal detachment, all of which responded transiently to systemic corticosteroids. Systemic cyclosporine was added, but the patient’s condition worsened to include fever, thrombocytopenia with both bleeding and bruising, hyperbilirubinemia, and elevated liver enzymes. Abdominal ultrasound revealed hepatic parenchymal lesions and bone marrow cytology showed hemophagocytosis and low NK-cell activity. Plasma EBV DNA levels were markedly elevated at 32100 IU/ml. The authors diagnosed presumed cyclosporine-induced immunosuppression and liver toxicity leading to fulminant EBV-induced HLH. The patient refused further therapy and died of multiple organ failure. The authors cited a previous report by Lee and Mieler of uveitis and exudative detachment in a patient with HLH.⁵¹ Others have described rare ocular involvement in HLH to include panuveitis⁵² and Purtscher-like retinopathy.⁵³

Bustos-Mejia et al³⁶ described a 68-year-old woman seen in a referral clinic in Bogota, Colombia, with necrotizing nodular scleritis complicated by perforation and requiring scleral graft placement who had elevated plasma anti-EBV IgG titers and in whom EBV DNA was amplified by PCR performed on a scleral biopsy taken from the graft site. Workup was otherwise unrevealing. The patient improved over three months following surgery while on topical corticosteroids and systemic valacylovir. A similar case of purported bilateral EBV-associated scleritis complicated by perforation in a 71-year-old Japanese woman with elevated serum anti-EBV antibody titers has been reported.⁵⁴ The authors suggested that the serologic and PCR-based test results, together with the slow therapeutic response to combined corticosteroids and an anti-viral agent, supported the causal association with EBV. Whether or not EBV did, in fact, cause or contribute to the necrotizing sclerokeratitis described in these two cases remains circumstantial and somewhat speculative, however.

Together, these studies add to the growing number of reports identifying EBV in patients with ocular inflammation. As noted, the most convincing evidence for a direct role of EBV comes from its identification in those eyes with lymphoproliferative disorders known, in some, to be induced by EBV, including B- and NK/T-cell lymphoid tumors. For most patients with ocular inflammation, however, the presence of either EBV DNA or anti-EBV antibodies appears to be both non-causal and of limited or no clinical utility. If, for some reason, ocular levels of EBV DNA or anti-EBV antibodies are assayed and found to be elevated, consideration should be given to the possibility of an intraocular EBV-associated malignancy and, independently, HIV co-infection.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.