Alternative Routes of Zoonotic Vaccinia Virus Transmission, Brazil

To the Editor: Vaccinia virus (VACV) causes exanthematous disease (bovine vaccinia) in Brazil. Outbreaks of this disease in humans have been reported since the late 1990s and have spread throughout Brazil (1). Natural human infections with VACV occur by close contact with infected cattle during milking. Lesions can spread to secondary body sites (forearms, arms, and face). Thus, personto-person transmission occurs (1). Moreover, virus can persist in household environments, remain infectious, and be transmitted by fomites (2). Although raw milk and cheese are potential sources of infection, no clinical cases have been associated with this transmission route (3,4). Data for person-to-person transmission in Brazil are scarce, but person-to-person transmission was recently reported (5). We report a possible case of person-to-person transmission of VACV. This study was approved by the Research Ethics Committee of Universidade Federal de Minas Gerais (registration protocol FR-413704). In September 2012, during a serologic survey in a rural area of Serro City (18°36′17′′S, 43°22′46′′W), Minas Gerais, Brazil (online Technical Appendix Figure, panel A, http://wwwnc.cdc.gov/EID/ article/21/12/14-1249-Techapp1.pdf), blood samples were obtained from a family of 5 persons (father, mother, and 3 daughters). The father and mother were 48 and 53 years of age, respectively, and had been vaccinated against smallpox. They reported contact with cows and horses (online Technical Appendix Table 1). Only the father had milked cows. The 3 daughters (13, 13, and 14 years of age) did not engage in any exposure activity. However, all family members had consumed raw milk and cheese. Bovine vaccinia lesions were observed on the hand of the father (online Technical Appendix Figure, panel B). In 2011, he had vesicular disease (no laboratory diagnosis) with clinical and epidemiologic features (lesions) suggestive of bovine vaccinia on his hands and forearms and systemic symptoms (fever, headache, malaise, myalgia, lymphadenopathy, and abdominal pain). His symptoms were mild and without any systemic clinical features. Two lesions developed on his hands and dried swab samples were collected from both lesions. Swab samples were processed as described (2) and used for virus isolation and molecular diagnosis. On the basis of previous studies that detected viral DNA in clinical samples from persons with bovine vaccinia (1), we used a quantitative PCR to amplify the vgf and ha genes of VACV (3–5), a standard PCR to detect the ha gene (3–5), and a seminested PCR to detect the ati gene (F.L. Assis, unpub. data). Serum samples were used for detection of virus-neutralizing antibodies (orthopoxvirus 50% plaque-reduction neutralization test) and molecular diagnostic studies (1). Virus isolation was attempted in Vero cells and chorioallantoic membrane. All results were negative. The 50% plaque-reduction neutralization test showed that the father, mother, and 14-year-old daughter had neutralizing antibodies against orthopoxvirus (titers 800, 3,200, and 800 neutralizing units/mL, respectively). All family members had positive results by molecular diagnostic test for >1 virus gene (online Technical Appendix Table 1). To rule out infection with parapopoxvirus, a complementary PCR (6) was also performed, and all family members had negative results. Quantitative PCR products for the ha gene from 3 virus-positive samples were sequenced in both directions in triplicate (Mega BACE Sequencer; GE Healthcare, Little Chalfont, UK). Sequences were aligned by using ClustalW (http://www.genome.jp/tools/clustalw/) and MEGA4.1 (http://www.megasoftware.net/) and showed 100% identity with each other (Figure). A phylogenetic tree was constructed by using the neighbor-joining method and 1,000-bootstrap replicates in the Tamura-3 parameter model (MEGA4.1). Sequences were grouped with VACV group 2 isolates. Sequences obtained were deposited in GenBank under accession nos. KP889223–5). In Brazil, outbreaks of bovine vaccinia are associated with rural environments. However, some clinical and