Sampling of herpes zoster skin lesion types and the impact on viral DNA detection.

This was a multicenter, non-therapeutic study to determine the optimal type of lesion sample for quantitative PCR detection of varicella zoster virus (VZV) DNA in herpes zoster patients. Up to three crusts, three crust swabs, three vesicle swabs, and three papule swabs were collected from 41 adults with clinically diagnosed herpes zoster. 83% of subjects had at least one valid crust swab (detectable VZV or β-actin DNA), 78% had at least one valid crust, 78% had at least one valid vesicle swab, and 32% had at least one valid papule swab. Of valid samples, 97% of crusts, 94% of vesicle swabs, 90% of crust swabs, and 84% of papule swabs were VZV-DNA-positive (≥10 DNA copies/sample). 37 (90%) subjects had at least one VZV DNA-positive sample. VZV DNA copy numbers were highest for vesicle swabs and crusts. The probability of a false-negative result was 5% for crusts, 6% for vesicle swabs, 14% for papule swabs, and 24% for crust swabs. Overall, vesicle swabs and crusts were the most specific and sensitive samples for detecting VZV.