Transmission of Varicella Vaccine Virus, Japan

To the Editor: Varicella-zoster virus (VZV), a human herpesvirus, is the causative agent of varicella (chick-enpox) and herpes zoster (shingles). Worldwide, children are routinely vaccinated with a live attenuated vari-cella vaccine containing the Oka vaccine (vOka) strain of VZV, originally developed in Japan (1–3). Although the risk for secondary transmission of the vOka strain from immunocom-promised vaccinees to susceptible persons is relatively high, the risk for transmission from immunocompetent vaccinees is low (1). We report secondary transmission of the vOka strain from an immunocompetent girl with a history of varicella vaccination to her healthy susceptible brother. Herpes zoster developed in a healthy 3-year-old girl 2 years after she had received the varicella vaccine (lot VZ040; Biken, Osaka, Japan). She received oral acyclovir treatment and fully recovered by day 19 after herpes zoster onset. On the same day that the girl recovered, her immunocompetent 2-year-old brother was found to have fever and a rash consisting of 10–20 papulovesicles; mild varicella was diagnosed. The boy had no known history of contact with persons infected with varicella or with persons who administered the varicella vaccine. After receiving oral acyclovir treatment, the boy recovered without systemic complications. On day 19 after the girl's onset of herpes zoster, an enzyme immunoas-say (Denka Seiken, Tokyo, Japan) confirmed the presence of VZV-specific immunoglobulin (Ig) G (titer 48.9, well above the detection limit of 2.0) but not IgM. The boy showed serocon-version of VZV-specific IgG from a ti-ter of <2.0 on day 3 after his disease onset to 19.3 on day 30. Although ve-sicular fluid or crust specimens were obtained from both children, only the specimens from the boy contained detectable amounts of VZV DNA. To determine whether vOka or a wild-type VZV strain caused the va-ricella in this boy, we performed PCR to amplify the entire region of gene 62 and determine its sequence, as described previously (4). The DNA sequence of the PCR product matched that of the vOka sequence with the exception of a single wild-type nucleotide substitution at position 105705 (Figure, GenBank accession no. AB497598). Restriction fragment length polymorphism (RFLP) analysis of the PCR products of the open reading frame (ORF) 38 and ORF54 loci using PstI and BgII (5) demonstrated that the strain had a vOka-like pattern , i.e., PstI-BgII+. Furthermore, the vOka-specific sequences at positions 5,745 and 94,167 were conserved in the strain. Taken together, these results indicate that the strain in …