Despite the brief surveillance period and unanticipated logistical impediments encountered in our study, over 60% of the patients that we sampled in two ecologically distinct regions of Guinea had evidence of arboviral syndromes. Five distinct arbovirus infections were noted. With the exception of yellow fever, these are the first reported cases of human disease caused by these viruses in Guinea to our knowledge. These results strongly suggest that various arboviruses circulate, are common causes of human disease in Guinea, and are likely to be markedly underdiagnosed.
Perhaps the least surprising finding of our study is identification of cases of YFV as YFV is endemic in sub-Saharan Africa.10
Despite a safe and effective vaccine, cases of YFV infection were reported in 13 of 14 West African countries at risk from 2000 to 2006, including numerous outbreaks in Guinea (), although there was no recognized outbreak of YFV or any other arbovirus in Guinea during the period of our study.10,11,30,31
It should be noted that only two of the subjects in our study reported having received the YFV vaccine—one CHIKV case from FRH and one case who was negative for all viruses tested. Ten subjects reported never receiving YFV vaccine. Thirty-five subjects did not know their vaccination status, including the two positive for anti-YFV IgM antibody. The resurgence of YFV in West Africa is partially attributable to the end of the routine preventative immunization campaigns common during the colonial era, which ended in the late 1950s/early 1960s; however, many African countries, including Guinea, have added YFV vaccine to the recommended list of routine childhood vaccines.30
Our finding of two cases of YFV during such a brief period of more intensive surveillance suggests that YFV infection may be more common that typically noted or assumed. We suspect that more thorough surveillance, at least in rural areas, would show a baseline incidence of YFV infection that is > 0. Such a finding would have significant consequences given the present policy of implementing costly and labor-intensive mass vaccination campaigns based on the confirmation of a single case. Declaration of an outbreak should ideally take into account the level of baseline transmission, but such an approach will require markedly enhancing surveillance and laboratory capacity.
CHIKV infection was frequent in our study (17% of cases), a finding in keeping with previous reports from Guinea showing a seroprevalence of CHIKV antibody over 50%13
(E. Jentes, unpublished data). There are also numerous reports of CHIKV isolations from various arthropods and small mammals in Guinea.12,15,32
The geographic distribution of CHIKV virus includes most of sub-Saharan Africa, although the virus has not frequently been the subject of study in West Africa.33–36
CHIKV is considered endemic in most rural areas, with small numbers of cases occurring each year, whereas large and explosive periodic outbreaks may occur in urban areas.13,34,37–40
The finding of anti-CHIKV antibodies in 30–100% of the population in some studies suggests asymptomatic or mild infection to be frequent, although underdeveloped surveillance systems to detect and diagnose cases most certainly contribute to this finding and interpretation.34,35
We found one confirmed case of DENV in Guinea, the first human case to be reported from that country, although DENV-2 was isolated from a mosquito in 1996.12
The volume of serum was insufficient to perform further testing to identify the specific infecting DENV serotype. Further evidence that DENV circulates in Guinea comes from a 1-month pilot study that we conducted in N'Zérékoré in 2004 in which 3 of 13 (23%) febrile patients tested positive for anti-DENV IgM antibody, whereas IgG antibody to DENV was found in 34 of 261 (13%) convenience samples using the PanBio assays. Furthermore, only two of eight (25%) persons in the study known to be previously vaccinated against YFV tested IgG antibody positive for DENV, suggesting that YFV cross-reaction did not account for the majority of the positive results.
DENV-1, -2, and -4 have been repeatedly isolated from humans and mosquitoes in various countries of West Africa, including Nigeria, Senegal, and Côte d'Ivoire, although DENV-3 has been recently reported in Cape Verde and isolated in travelers returning from Senegal and Côte d'Ivoire.41–51
Reports of cases of dengue fever in Africa have increased over the past two decades, although the dramatic epidemics seen in other parts of the world have not occurred.52,53
All four serotypes have been found in sub-Saharan Africa, although dengue hemorrhagic fever has only been reported in a few isolated cases.52,53
Recently, it has been proposed that African sylvatic DENV strains are less virulent than strains circulating in other parts of the world and therefore, would not cause severe human disease.54,55
WNV was the most common arbovirus infection noted in our study. WNV was reportedly isolated from a wild rodent in Guinea in 2006, but this is the first report of human infection.32
Serological evidence of WNV infection in humans, horses, birds, and arthropods has been reported in neighboring Senegal and Cote d'Ivoire as well as in one confirmed case in a traveler returning from Senegal.56–60
TAHV is widely distributed in Europe and Asia but has not been thoroughly investigated in Africa. Serological evidence of human infection in Cameroon and small- mammal infection in Tunisia62
has been reported.61
To our knowledge, this is the first report of symptomatic TAHV infection in Africa.
Our study had several limitations. The various impediments imposed by the civil unrest in Guinea are described above. The possibility that the blood samples could contain Lassa virus necessitated heat inactivation, precluding any attempts at virus isolation, the gold standard and definitive evidence of arbovirus infection, or polymerase chain reaction (PCR). Frequent power outages posed a challenge to maintaining the cold chain, properly storing specimens and reagents, and performing the onsite serological tests. This perhaps accounts for the discrepancy between the negative findings in Guinea and the frequent positives found at CDC. Furthermore, the circuitous shipment of the serum specimens to the United States through Sierra Leone and Cote d'Ivoire resulted in numerous breaks in the cold chain, which ultimately precluded all but serological diagnosis and could have even resulted in deterioration of antibodies measured through these tests. However, if that were the case, our results represent an underestimate of the incidence of arbovirus infections in Guinea. It must also be noted that cross-reactions are always a potential issue in the serological diagnosis of arbovirus infections, although the conservative interpretation of the PRNTs performed in this study should limit any undue conclusions regarding the specific virus infections. Furthermore, even if cross-reactions result in errors identifying the specific infecting virus, the important conclusion that many arboviruses are circulating in the area remains valid. Lastly, we were not able to make detailed clinical observations of the patients, test for coinfections, or completely exclude the possibility of drug-resistant malarial or bacterial infections.
The results of our study should serve as a reminder to clinicians in Guinea that arboviruses are a frequent cause of febrile disease in that country and likely, the rest of West Africa as well. More intensive surveillance systems are needed to define the scope of the problem, but it can safely be assumed to be large. Improving laboratory capacity in sub-Saharan Africa for the diagnosis of arbovirus infections will be integral to a full understanding of the burden posed by these agents.