Article Text

Download PDFPDF

Herpes simplex virus in the human cornea
  1. University of Nottingham, University Hospital, Queen's Medical Centre, Nottingham, NH7 2UH
  1. harminder.dua{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Herpes simplex virus (HSV) infection is prevalent throughout the world. An estimated 60–70% of children aged 5 and close to 90% of adults exhibit seropositivity (neutralising antibody) to HSV antigens. Despite this widespread prevalence, only 20%–30% present with clinical disease, and ocular manifestations develop in less than 1%.1

Traditionally, the diagnosis of herpes simplex keratitis (HSK) has been based on clinical evaluation, occasionally complemented by viral culture. Immunohistology and, relatively recently, polymerase chain reaction (PCR) and in situ hybridisation (ISH) techniques have been employed to detect viral DNA in corneal buttons removed at the time of penetrating keratoplasty or in corneal biopsy specimen taken from patients with undetermined corneal inflammation.2-7 The deployment of these techniques, however, has largely been as part of studies or projects, rather than a routine clinical service. The sensitivity and specificity of these techniques is variable and the detection of viral DNA does not necessarily correspond with presence of infective virus, as only a part of the viral DNA may be present. Only viral culture is 100% specific, but its sensitivity is rather low. Like many diagnostic tests, these tests are not mutually exclusive but complementary. In the paper by Kaye et al in this issue of the BJO (p 563) the authors have compared the sensitivity and specificity of viral culture, immunohistology, PCR, and ISH in detecting virus in 110 human corneal buttons removed at penetrating keratoplasty and 19 eye bank donor eyes not suitable for corneal transplantation. They concluded that both PCR and immunohistology were sensitive for detection of HSV-1 in the human cornea and, when combined, the specificity for the diagnosis of HSK reached 97%. The clinical definition of HSK, as a history of recurrent dendritic or geographic corneal ulceration and the development of stromal scarring, was used as a standard to measure sensitivity and specificity.

A large number of studies on human3-6 8 9 and animal10 11 corneal buttons, obtained during the quiescent stage of HSK, have reported the detection of viral DNA using PCR and immunohistology (ISH was not found to be as sensitive8). Holbach et al 6 were able to detect viral DNA with equal frequency from vascularised and non-vascularised HSK corneas but noted a significantly higher detection of viral antigen, by immunohistology, in the avascular specimen. These studies show that the yield of viral DNA from both animal and human corneal tissue obtained from quiescent post-HSK eyes tends to decline with time. The longer the duration following the active episode, the less likely it is to detect viral DNA. This is probably related to the decline in the amount of viral DNA present. Interestingly, all the above studies, like the one reported by Kaye et al in this issue, also reported detection of viral DNA from non-HSK corneal buttons. The detection rate was not as high but was nevertheless significant. Furthermore, several studies have shown the presence of HSV DNA in normal and eye bank donor corneas.3 12-16 Primary graft failure and massive loss of endothelium in stored eye bank corneas, has been associated with viral DNA in the affected corneal tissue.12 17

The presence of viral DNA in corneas, long after the active HSK episode has subsided, has led to the hypothesis of extraneuronal herpetic latency in the cornea. This concept is rapidly gaining popularity but is not yet totally convincing. The presence of viral DNA does not correspond with viral infectivity of the tissue. This has been borne out both in cell culture studies and in vivo, after transplantation of viral DNA positive corneas to non-herpetic recipients. Morris et al 13analysed the spent culture media of 80 donor corneas and detected HSV DNA in three cell pellets; however, follow up of patients receiving these corneas did not reveal HSV eye disease or graft failure. In one study in rabbits, aimed at establishing HSV latency, only 10% of eyes contained virus that could be reactivated in culture while an additional 55% contained viral DNA.10

What then is the take home message for the practising clinical ophthalmologist? The diagnosis of HSK is still essentially clinical. Although it is always reassuring to have laboratory confirmation of a clinical diagnosis, one must not lose sight of the fact that the standard against which the sensitivity and specificity of the above tests was measured, was the clinical definition (diagnosis) of HSK. The apparent “widespread” presence of viral DNA in HSV, non-HSV, and even normal corneas is cause for concern but the mere presence of viral DNA does not automatically equate to its ability to cause infection. Corneal latency, however real or otherwise that might be, is not the same as neuronal latency. The comparatively low levels of viral DNA which is transcriptionally dormant, the lack of adequate numbers of latency associated transcripts, the possibility of presence of incomplete viral genome and its attenuation with time in corneal tissue, all serve to diffuse any concern over transplanting “infected normal corneas” to healthy recipients, at least at the present state of our knowledge. One question remains. If the amount of viral DNA and its detection rate declines with time in HSV infected corneas, does its detection in donor corneas imply viral reactivation in the period immediately preceding death? If so, then should these not be potentially infective?