Epidemiology and molecular virus characterization of reemerging rabies, South Africa

Abstract

The incidence of dog rabies in Limpopo Province, South Africa, increased from 5 cases in 2004 to 100 in 2006. Human rabies had last been confirmed in 1981, but investigations instituted after an index case was recognized in February 2006 identified 21 confirmed, 4 probable, and 5 possible human cases between August 5, 2005, and December 31, 2006. Twelve of these case-patients were identified retrospectively because the diagnosis of rabies was not considered: 6 of these patients consulted a traditional healer, 6 had atypical manifestations with prominent abdominal symptoms, and 6 of 7 patients tested had elevated liver enzyme activity. Molecular genetic analysis indicated that outbreak virus strains were most closely related to recent canine strains from southern Zimbabwe. Delayed recognition of the human cases may have resulted from decreased clinical suspicion after many years of effective control of the disease and the occurrence of atypical clinical presentations.

Figure 1

Provinces of South Africa, and neighboring countries. Inset shows a choropleth map of the number of confirmed dog rabies cases by district in Limpopo Province in 2005–2006 and the location of human cases (4 case-patients, for whom coordinates of place of residence were unavailable, were excluded).

Figure 2

Laboratory-confirmed animal rabies cases, Limpopo Province, South Africa, 1994–2006.

Figure 3

Numbers of possible, probable, and confirmed human cases and laboratory-confirmed domestic dog rabies cases by month of diagnosis, Limpopo Province, South Africa, 2005–2006.

Figure 4

Neighbor-joining tree of canid rabies viruses from humans and animals from Limpopo (LP), Mpumalanga (MP), North West (NW), Free State (FS), Eastern Cape (EC), Northern Cape (NC), KwaZulu-Natal (KZN), and Western Cape (WC) Provinces of South Africa (SA) and neighboring countries of Zimbabwe, Mozambique, and Namibia. The Pasteur virus strain (PV) was used as the reference strain in the sequence alignment. Horizontal scales represent the evolutionary distance; vertical lines are for clarification purposes only. The scale bar indicates nucleotide substitutions per site. Viruses are identified by a laboratory reference number, source animal, locality of origin, and year of isolation. A–F represent virus lineages supported by bootstrap values of >70%; sublineages are indicated numerically. *Identical strains.