Objective: To determine the accuracy of three commercial enzyme immunoassays in detecting and subtyping antibodies to herpes simplex virus type 1 or 2.
Design: Cross-sectional.
Setting: Referral medical center.
Patients: Ninety patients with culture-positive lesions caused by infection with herpes simplex virus type 1 or 2. The results of Western blot and glycoprotein G immunodot enzyme assays showed that an additional 53 patients had subclinical herpes simplex virus 2 infection, that another 20 patients had subclinical herpes simplex virus type 1 infection, and that 23 patients were seronegative.
Measurements: Three commercial enzyme immunoassays were used to determine herpes simplex virus antibody subtypes.
Main results: All three commercial assays performed poorly in all patient groups (except in patients who were seronegative for herpes simplex virus). Among the 40 patients with a first episode of genital herpes, seroconversion to the appropriate viral type was shown by the three assays in only 33%, 55%, and 75% of cases. Among patients with recurrent genital herpes, the three commercial assays identified more than 90% of patients with only herpes simplex virus type 2 antibodies but failed to identify herpes simplex virus type 2 infections in 58% to 76% of patients with antibodies to both virus subtypes. The three assays correctly identified only 55%, 75%, and 85% of the 53 "silent carriers" of herpes simplex virus type 2. Overall, the three enzyme immunoassays detected herpes simplex virus type 2 antibodies in 60%, 62%, and 93% of patients with subtype 2 infections and falsely detected type 2 antibodies in 8%, 27%, and 49% of patients with type 1 infections.
Conclusion: Currently licensed enzyme immunoassays give inaccurate or misleading results about the correct herpes simplex virus infecting subtype.