Endemic human monkeypox, Democratic Republic of Congo, 2001-2004

Abstract

By analyzing vesicle fluids and crusted scabs from 136 persons with suspected monkeypox, we identified 51 cases of monkeypox by PCR, sequenced the hemagglutinin gene, and confirmed 94% of cases by virus culture. PCR demonstrated chickenpox in 61 patients. Coinfection with both viruses was found in 1 additional patient.

Figure 1

Distribution of 52 confirmed cases of human monkeypox (MPX) by health zone in the Democratic Republic of Congo (DRC), 2001–2004. The cumulative number of cases per province is in parentheses. Confirmed cases of MPX originated from a total of 26 different health zones located in 5 provinces of DRC; most (70.5%) were reported from Kasai Oriental and Equateur Provinces. Two groups can be differentiated on the basis of sequence of the hemagglutinin gene: light gray, group 1; dark gray, group 2. Note: The boundaries of the DRC health zones have since been redrawn. Although the health zones Tshopo and Mweka are not shown (located in provinces Orientale and Kasai Occidental, respectively), the general area is highlighted to represent MPX cases in these regions.

Figure 2

Phylogenetic inference relationships of the open reading frames encoding the hemagglutinin protein of selected strains of vaccinia, variola, cowpox, and monkeypox viruses and monkeypox virus isolates described in this study. Sequences used were cowpox virus Brighton (AF375089), variola virus Bangladesh (AF375129), vaccinia virus Lister (AY678276), monkeypox virus mpv 1997 (AF375096), mvp-squir (AF375112), mpv Zaire 77-0666 (Z99052), mpv-cncr (AF375102), mpv 74-226 (AF375099), mpv-082 (AF375095), mpv-utc (AF375113), and mpv-3945 (AF375098). ClustalW, version 1.83 (10), was used to generate amino acid multiple sequence alignments (pairwise gap opening = 35 and gap extension = 0.75; multiple alignment gap opening = 15 and gap extension = 0.30; Gonnet series). Each alignment was processed using RevTrans (11). Bayesian posterior probability inference of phylogeny used MrBayes, version 3.084. MrBayes settings for the best-fit model (GTR+I+G) were selected by hierarchies for the likelihood ratio test in MrModeltest 2.0 (12). Bayesian analysis was performed with MrBayes; the maximum likelihood model used 6 substitution types (nst = 6). Rate variation across sites was modeled by using a gamma distribution, with a proportion of sites being invariant (rates = invgamma). The Markov chain Monte Carlo search was run for 1 million generations; trees were sampled every 100 generations (the first 4,000 trees were discarded as burn-in). NL, the Netherlands; DRC, Democratic Republic of Congo.