During August 2009, a United Nations Food and Agriculture Organization mission was undertaken with local counterparts in Congo to support development of an action plan to control ASF. A key factor in selecting the sites sampled was inclusion of the main pig-producing, marketing, and consuming areas, with a particular focus where suspected ASF outbreaks had been recently reported. From the survey conducted, 86 samples comprising serum (35 samples), whole blood (44 samples), and tissues (7 samples) were collected from 80 domestic pigs in the departments of Brazzaville, Pointe Noire, Kouilou, Bouenza, Niari, and Cuvette (Figure). These departments contain 80%–90% of the country’s pig population. Samples also were collected from Pool in southeastern Congo, where a recent hemorrhagic disease outbreak, characterized by case-fatality rates of ≈80%, had been reported. Clinical material was sent to the European Union Reference Laboratory for African Swine Fever (Centro de Investigación en Sanidad Animal, Madrid, Spain) for confirmatory diagnosis and characterization of the ASFV strain(s) responsible for the outbreak(s).

Specific ASF antibodies were detected by using the OIE-prescribed assays (8) in 7 of 35 serum samples analyzed. All positive serum samples were from animals that had survived the ASF outbreaks in Bouenza, Niari, and Pointe Noire during 2008. For ASFV genome detection, OIE-prescribed PCRs (8) were performed on DNA extracted from 28 serum samples, 44 blood samples, and 7 tissue homogenates. A positive result was obtained in 17 (21%) of samples analyzed, indicating ASF in all the departments where the survey was conducted except for Kouilou.

Subsequently, 5 hemadsorbent Congo ASFVs (Table) were isolated in porcine peripheral blood macrophages (8). Viral DNA was extracted, and 3 different sets of primers were used for ASFV genotyping. A region of 478 bp at the C-terminal end of the p72 protein and the full-length sequence of p54 gene were amplified by using primers p72U/D and 89/722, respectively (6). We compared the sequence from each of the p72 and p54 amplicons with homologous sequences representative of each previously described p72 and p54 genotype (5,6). A rooted minimum-evolution tree was constructed by using MEGA4.0 software and the p-distance nucleotide substitution model (9). Two different ASFV phylogenetic profiles were found in western and southeastern Congo. The phylogeny inferred for the ASFV isolates from eastern Congo (Con09/Pk45, Con09/Bzz020, Con09/Abo) placed them within p72 and p54 genotype IX with viruses collected from recent outbreaks in Kenya and Uganda during 2006–2007 (6,7). By contrast, ASFV isolates from western Congo (Con09/Ni16, Con09/PN003) clustered in p72 genotype I and p54 genotype Ic (6) comprising historical isolates from western Africa, such as Ang72 and Kat67, and the first ASFV isolated from Europe, the Portuguese Lisbon57 isolate (Figure A1).

To provide higher resolution, the tetrameric tandem amino acid repeat sequences within central variable region (CVR) of the B602L gene were analyzed (5–7,10). Primers CVR1 (5′-ACTTTGAAACAGGAAAC(AT)AATGATG-3′) and CVR2 (5′-ATATTTTGTAATATGTGGGCTGCTG-3′) were designed in this study selected from complete genome sequences of ASFV strains available in GenBank.

Under the same thermal cycling conditions used for full p54 gene amplification, amplicons of 600–650 bp were generated from the isolates from eastern Congo, whereas the estimated size of the amplicons from western Congo was 500 bp (data not shown). Analysis of the tandem amino acid repeat sequences within the CVR demonstrated 3 different variants cocirculating in the country, 2 within the p72 type I virus genotype. The type of CVR sequence (AABNABTDBNAAAA) identified in the isolate from Pointe Noire (Con09/PN003) was identical to CVR subgroup XIII (11) that contains early isolates from Angola (Ang70, Ang72) and Portugal (Lis57). However, the Con09/Ni16, also classified within p72 and p54 genotypes I and Ic, showed a unique CVR sequence (AAAAAAAAAF) not previously described but most similar to viruses from Burundi and Kenya in CVR subgroup XXVI (7,11). As for p72 and p54 genotyping, isolates from eastern Congo were related to CVR subgroup XXIV, which contains isolates obtained during the 2006 and 2007 outbreaks in Uganda and Kenya (6,7).

We confirmed ASF in 5 of the 6 departments in which surveillance was conducted during August 2009 in Congo. Genotyping of 5 ASFV isolates from Congo resulted in identification of genetically distinct viruses circulating simultaneously in the country. In eastern districts of Congo, viruses were most genetically similar to those recovered from the outbreaks in Kenya and Uganda during 2006 and 2007. Therefore, genotype IX, associated with ASF outbreaks in eastern Africa during the previous 10 years, also occurs in western Africa. By contrast, viruses in western Congo were similar to genotype I viruses that historically circulated in western Africa. Analyses of the CVR identified 2 discrete variants within genotype I. The Pointe Noire ASFV was identical to viruses isolated in neighboring Angola during the 1970s, indicating a prolonged persistence of this virus type in the region. By contrast, the ASFV isolated from Niari was related to variants circulating during the 1980s in Burundi. These results are consistent with recent transfer and dissemination of ASFV genotypes from eastern to western Africa. Broadening of our initial findings requires extending analyses to western-central Africa, particularly to countries such as Cameroon and the Democratic Republic of the Congo, and providing more in-depth analysis in Congo.

The apparent recent movement of a virus genotype previously associated with virulent ASF in eastern Africa to western Africa, where the viruses have hitherto always been classified in p72 genotype I, indicates the possibility of future outbreaks of disease caused by novel viruses in western Africa. The reality of this scenario recently was demonstrated in the Caucasus region. Our study confirms the continuing spread of ASFV.