Comparison of Immunocytology to Tissue Culture for Diagnosis of Presumed Herpesvirus Dendritic Epithelial Keratitis

Objective: The objective of this study is to prospectively compare the sensitivity and specificity of immunodetection of herpes simplex virus (HSV) in impression cytology specimens obtained directly from presumed herpesvirus dendritic epithelial keratitis with virus isolation by tissue culture of cells scraped from the same lesion. Methods: Corneal impression cytology and tissue culture were performed on 29 consecutive patients presenting with presumed herpesvirus dendritic epithelial keratitis during a 6-month period. Impression cytology of dendritic epithelial keratitis lesions with Millipore Biopore membranes were evaluated for the presence of antigens specific to HSV type I (HSV-1), HSV-2, and varicella-zoster virus (VZV) using monoclonal antibodies specific to these herpesviruses and immunofluorescent staining techniques. Results: Tissue culture was positive for HSV-1 in 52% (13 of 25) of dendritic epithelial keratitis patients without skin lesions, and was negative for VZV in 4 patients with dendritic epithelial keratitis and skin lesions in the distribution of the first division of the trigeminal nerve. The remaining 12 tissue cultures showed no cytopathic effect. Compared with tissue culture, impression cytology was 100% sensitive (13 of 13) and 92% specific (11 of 12) for the diagnosis of HSV-1 dendritic epithelial keratitis (Kappa coefficient of agreement 0.92). Although our sample size for VZV dendritic epithelial keratitis was small, the impression cytology findings correlated with our clinical diagnosis more often than tissue culture (2 of 4 versus 0 of 4). Conclusion: Impression cytology allows simultaneous debridement of dendritic epithelial keratitis and, when combined with immunocytologic staining procedures, provides a simpler, more rapid, and less expensive alternative to tissue culture for the diagnosis of dendritic epithelial keratitis caused by HSV or VZV.