Detection of herpes simplex virus DNA from genital lesions by in situ hybridization

Lesion specimens from 118 episodes of recurrent genital herpes were used to compare herpes simplex virus (HSV) isolation with a direct specimen test for in situ DNA hybridization utilizing a biotinylated probe. The frequency of detection of HSV was similar with both tests; HSV was isolated from 81% of vesicular lesions, 76% of pustules, and 67% of ulcers, while HSV DNA was detected in 77, 76, and 55% of lesions in these stages, respectively. Utilizing both methods, HSV was identified in 91, 94, and 79%, respectively. The sensitivity and specificity of the DNA probe in comparison to standard viral isolation in tissue culture were 92 and 63%, respectively. Seven DNA-positive, viral isolation-negative specimens were obtained from patients who had positive culture confirmation at some time subsequent or prior to enrollment, suggesting that these were true positive results. The sensitivity of the DNA probe was dependent on cellular content of the specimen, and 36 (28%) of the 127 submitted specimens had fewer than 20 nonsuperficial cells. The DNA probe was rapid and convenient; its major disadvantage was the lack of type-specific information. The performance of the probe in lower-prevalence populations and in asymptomatic shedding of HSV remains to be evaluated.