Chikungunya outbreak in a rural area of Western Cameroon in 2006: A retrospective serological and entomological survey

Chikungunya virus (CHIKV), is a member of the Alphavirus genus (Togaviridae family) and belongs to the Semliki virus antigenic complex comprising species with similar clinical manifestations and close antigenic patterns like CHIKV, O'Nyong-nyong virus (ONNV), Ross River virus and Mayaro virus [1, 2].

CHIKV is now known to circulate in enzootic cycles throughout much of sub-Saharan Africa where it originated. In West and Central Africa, the disease is maintained in a sylvatic cycle involving wild non-human primates and forest-dwelling Aedes spp. mosquitoes. The virus has been isolated from sylvatic mosquito species in several countries including Senegal, Côte d'Ivoire, Central African Republic and South Africa. The mosquito species involved vary geographically and with ecological conditions; however, the major species involved in sylvatic cycles are Ae. furcifer, Ae. taylori, Ae. luteocephalus, Ae. africanus and Ae. neoafricanus [3–5]. In rural regions, outbreaks of human disease are occasionally detected when adequate laboratory diagnostics are implemented. These outbreaks tend to be of small scale and appear to be heavily dependent upon the sylvatic mosquito densities that increase with periods of heavy rainfall [6].

In Asian and African urban cycles, the disease is transmitted by Ae. aegypti and Ae. albopictus with humans and mosquitoes as the only hosts of the virus [7, 8]. CHIKV has been responsible for many major outbreaks since its discovery in 1953 [9] but also has apparently disappeared for long periods during which it probably occurred only as sporadic or asymptomatic cases. From 2004, cases of CHIKV have risen meteorically in urban areas [10–13] leading to the recognition of CHIKV as a dangerous and important emerging arbovirus [14].

In Cameroon, cases of CHIKV were reported in 2006 in the two main cities of the country: Douala and Yaoundé [15]. That same year, in a rural locality in Kumbo (Bui division, Northwest Region of Cameroon), situated about 300 km north of Yaoundé, about 400 inhabitants of three neighbouring villages exhibited a dengue-like syndrome. From the clinical investigations carried out from October 31 to November 1, 2006 by the local health authorities, it appears that these clinical manifestations started in April 2006, at the beginning of the rainy season, becoming more serious in August. They were characterized by fever, chills, headache and muscle pains, combined with prolonged and severe joint pains. Unfortunately, this information reached the Ministry of Health (MOH) only in 2007 when a CHIKV epidemic was reported in Gabon, a neighbouring country [13]. The MOH then initiated field investigations so as to identify the aetiology of this fever and to define its extent in the area.

This paper describes the findings of the serological and entomological surveys conducted from November 18-26, 2007, at the beginning of the dry season in three villages of Kumbo. Results of such a study are invaluable for public health systems in Central African region where epidemics are frequently underreported or misreported.

A total of 105 persons (sex ratio m/f 0.44) were sampled for blood collection of whom 95 reported suffering clinical manifestations of a dengue-like syndrome during the outbreak period (about one year before this survey). The mean age was 50 years (minimum 8, maximum 81) with a standard deviation of 17.5. Most of the people examined were farmers (85.8%).

Nevertheless, all the 95 sera (90.5%) that were positive for anti-CHIKV IgG antibodies were also positive for anti-ONNV IgG antibodies due to the cross reactivity occurring between CHIKV and ONNV. No antibodies to DENV were detected.

IgM seropositivity to CHIKV was similar between females (52%) and males (50%) (Table 1). Because few participants between 5 and 29 years were enrolled in the survey, 5-14 and 15-29 year age groups were combined for statistical purposes; thus, anti-CHIKV IgM antibody prevalence did not vary significantly across different age strata, neither did anti-CHIKV IgG antibodies.

During the entomological survey 78 householders gave their oral consent for inspection of their houses. Most of them reported suffering the acute febrile syndrome the previous year. Eighteen to 30 households of 3 to 6 houses were surveyed in each village. All water containers, in or around human dwellings were inspected for the presence of mosquito larvae. Culicine larvae were found in 4 peridomestic breeding sites (old buckets, puddles) outside human dwellings and in 3 locations situated in raffia bushes where water had collected. Out of 305 larvae collected in these different breeding sites, 87 developed to the adult stage; 56 (64.4%) were Aedes africanus and the remaining Culex spp. Moreover, a number of adults of Ae. africanus mosquitoes were identified biting in raffia bushes during the day and this was consistent with complaints by the local populations of a high burden of mosquito bites in this particular biotope. Although only 3 breeding sites were found with culicine larvae in this environment, it was evident that places where water collected among raffia bushes probably represented important breeding places for Ae. africanus immature stages.

The prevalence of anti-CHIKV IgM antibodies one year after the outbreak was high and similar to data recorded during previous outbreaks of CHIKV [16, 17]. Nevertheless, such a high prevalence was unexpected due to the delay between the outbreak and our study but was in accordance with the occurrence of sporadic cases of CHIKV infection in the area. Generally, high IgM antibody prevalences are recorded during the peak incidence of an outbreak as observed in the Comoro Island [17], but tend to decrease with time as reported by Sergon et al. on the Kenyan Island of Lamu [18] and Thonnon et al. [16] in Senegal during a yellow fever outbreak survey.

Although we did not collect blood samples from unaffected areas (control) for comparison purposes, the possibility of the persistence of IgM anti-CHIKV for one year cannot be ruled out. This apparent persistence of anti-CHIKV IgM is consistent with the observations of both Fourie & Morrison and Gauzere [19, 20] who described such persistence of IgM antibodies for up to 12 months in patients suffering from polyarthritis or rheumatoid arthritis after CHIKV infection; and with those of Olivier et al. [21] who recorded 75 and 42% of anti-CHIKV IgM positives, respectively after 7 and 10 months, in French patients with imported CHIKV between 2005 and 2007. These observations closely match the situation in our study where many people were still complaining of recurrent arthralgia (joint pains). However, as we did not include a control village in our study, we could not totally discard the possibility of endemic circulation of the virus in the area. The fact that 39% of the samples were only IgG anti-CHIKV antibody positive could also be a result of the herd immunity from previous exposure to the virus as observed in healthy rural populations of Southern Cameroon by Kuniholm et al. [22].

Although arboviral infections including CHIKV are common in sub-Saharan Africa [1, 13, 16, 23–26], data on their circulation and prevalence in Central Africa are poorly documented due to confusion between arboviral infection and hyperendemic Plasmodium falciparum infection and/or lack of diagnostic tools in local health care centres. Thus only a few sporadic cases of arboviral infections have been reported in recent years across Cameroon [15, 22, 27]. However, and despite evidence of the circulation of CHIKV in the country, this study to the best of our knowledge is the first report of an important rural epidemic of CHIKV in Cameroon.

The outbreak affected equally the three villages in Kumbo (Ngehndzen, Ndzeru and Tasaï), and both sexes, and all age groups. Despite the over-representation of women (sex-ratio 0.44) among the persons sampled during this survey, both sexes exhibited the same seropositivity rates. The possible explanation to this is that women, men and children are equally exposed to Aedes spp bites either in their dwellings or in the raffia bushes where they fetch firewood (women), collect palm wine (men) and water (children).

Despite the fact that the extensive serologic overlap occurring between CHIKV and ONNV precludes any definitive interpretation of the results as to which virus may have been responsible for the production of the IgG antibodies in 90.5% of the sera [28], the overall optical density of the CHIKV ELISA was three to eleven times more than that of ONNV (data not shown). This result is consistent with the comparative studies between CHIKV and ONNV supporting the one-way antigenic relationship between CHIKV and ONNV [29, 30].

As the CHIKV outbreak in Kumbo and Yaoundé/Douala occurred during the same period, it is possible that the same virus strain was implicated. It has been shown that the Cameroon strain is highly similar to the CHIKV strain reported in the sub-region including Democratic Republic of Congo and Gabon [15]. However, the occurrence during these recent years of epidemic or sporadic cases highlights the possible mutation of the virus, with the introduction of a new and competent vector Ae. albopictus into Africa [31]. Indeed isolates from the Indian Ocean (La Reunion, etc.) were found to carry a mutation at residue 226 of the membrane fusion glycoprotein E1 (E1-A226 V); and it was demonstrated that such a mutation provided a selective advantage for the replication and transmission of CHIKV by this mosquito [32, 33]. Ae. albopictus was identified as the primary vector in the Chikungunya outbreak in Libreville (Gabon) [34, 35] and was suspected in the transmission of CHIKV in Yaoundé and Douala [15]. This invasive species burst into the African continent in the early 1990s and is now present in most Central African countries [36–38]. However no Ae. albopictus or Ae. aegypti, were found during this survey probably because the study took place during the dry season which is probably not suitable for their emergence. Moreover, these two species were described in the North-West regional capital (about a hundred km to the south) during the rainy season [38].

Although we did not search for the virus in mosquitoes collected, Ae. africanus could at least be involved in the transmission of CHIKV down the chain of sylvatic/domestic mosquito species in this area. It is possible that by entering areas like raffia bushes, and so being in contact with the sylvatic Ae. Africanus, humans might have become infected therefore serving as their incidental hosts. These humans might have then provided a source of the virus to infect potential peridomestic mosquitoes (although this was not found during our survey), which then became involved in the transmission cycle of the virus. CHIKV is known to circulate in a natural cycle amongst forest-associated simians and sylvatic mosquito species in West and Central African jungles, spasmodically causing outbreaks of varying size and intensity amongst the local populations living near the jungles [39, 40].

Between April and July 2006, CHIKV was identified during an outbreak of a febrile syndrome in Yaoundé and Douala [15]. In the mean time, in a rural area of Western Cameroon, several people also suffered from acute febrile illness. Although investigations were carried out one year after the latter outbreak, these findings suggest a recent circulation of CHIKV in three villages of Kumbo (Western Cameroon). Ae. africanus which was found to be relatively abundant in the raffia bushes could be involved in the transmission of CHIKV along the chain of sylvatic/domestic mosquito species in this rural area via humans and simians that come into contact with the virus either in the gallery forest or raffia bushes surrounding the villages.

This study, alongside previous ones, provides further evidence of the circulation of the CHIKV in the Cameroon forest region. Despite the co-endemicity of malaria and arbovirus infections in the Central African sub-region CHIKV infection is an emerging public health threat which should be given particular attention. In the absence of a vaccine, more effort should be put into prevention by involving the local population in vector control activities.

MD is a young virologist in charge of the newly created Arbovirus and Haemorrhagic Fever virus branch in the Laboratory of Virology at Centre Pasteur Cameroon. Since 2006, his research interests have focused on diagnosis and epidemiology of arbovirus diseases (dengue, chikungunya, yellow fever, etc.).