been linked with FHF in neonates, pregnancy, immunocomMembers of the herpes virus family and hepatitis B promised patients, and apparently healthy individuals. [5][6][7][8] Epvirus (HBV) have been implicated as etiologic agents in stein-Barr virus (EBV) infection may cause acute liver failure non-A, non-B (NANB) fulminant hepatic failure (FHF), in male infants with X-linked lymphoproliferative disease, but the frequency of infection with these agents has not immunocompromised patients, and those in whom it is acbeen established using appropriate controls. To examine quired sporadically. these findings. 4,18,19 To more accurately define the contribuand 14 of 104 control patients (13%) were positive for tion of the herpes viruses and cryptic HBV infection to NANB CMV DNA. Two of 50 FHF (4%) and 10 of 104 control FHF, we systematically assessed the livers of both patients patients (10%) had EBV DNA, and HBV DNA was obwith acute liver failure and representative controls for eviserved in 3 of 10 North American FHF patients (30%) dence of viral sequences using the polymerase chain reaction and 3 of 59 controls (5%) without serum markers for HBV (PCR). tion was diagnosed by evaluating serum for anti-hepatitis A virus Received March 8, 1996; accepted August 9, 1996. immunoglobulin M, hepatitis B surface antigen and hepatitis B core immunoglobulin G (CMV IMX, Abbott Laboratories, Chicago, IL) and

Antigenic analysis using monoclonal antibodies and genomic analysis using ribonuclease protection was done on 47 isolates of group A respiratory syncytial virus (RSV) recovered from children in St. Louis during four RSV seasons. Antigenic analysis identified four subgroups; of the three that included more than one member, those designated A/2 and A/2V had characteristic ribonuclease protection patterns. A third subgroup, A/4, exhibited more extensive genomic heterogeneity, but all isolates were distinguishable from those in subgroups A/2 and A/2V. Individual RSV epidemic seasons included isolates representing multiple subgroups of group A and multiple intrasubgroup variants, in addition to isolates from group B. Isolates that were indistinguishable by either antigenic or genomic analysis were present in more than one epidemic season. The subgroups may represent parallel evolutionary lineages, whose relevance to RSV immunity and pathogenesis requires further study.

Several clinical varicella-zoster virus isolates obtained during testing of a live varicella vaccine had DNA restriction fragment patterns resembling neither vaccine nor wild-type virus [Gelb et al., J Infect. Dis. 155, 633-640, 1987]. One explanation for these isolates was recombination in vivo. To determine if such recombination is likely, two strains of varicella-zoster virus, distinguishable by restriction endonuclease fragment size differences (wild-type strain EF and the OKA vaccine strain), were grown together in tissue culture. After three passages, the mixed infection virus was plaque-purified. DNA from about 13% of the plaque-purified isolates had one or more BglI fragments found in neither parental virus. Hybridization studies showed that isolates containing one of the new BglI fragments were recombinants of the two parental strains. The BglI restriction fragment pattern of these recombinants resembled those of the unusual varicella isolates from individuals either vaccinated with the live attenuated OKA varicella vaccine and later exposed to natural varicella, or simultaneously exposed to both a recent recipient of the vaccine and natural varicella.

The DNA from several clinical isolates of varicella-zoster virus (VZV) were compared with the DNA from the vaccine strain VZV using three restriction endonucleases: BamHI, BgII, and HpaI. When electrophoresed through an agarose gel, the vaccine DNA digestion pattern was significantly different from the digestion patterns of the wild-type DNAs. Variations in the digestion pattern of the separate clinical isolates were also observed.