Zika Virus in West Africa: A Seroepidemiological Study between 2007 and 2012

According to the World Health Organization, the entire African continent is at risk of a Zika outbreak. To increase data availability on the epidemiology of Zika virus circulation in Africa, we evaluated the immunity to Zika virus in a selected cohort of subjects from West Africa between 2007 and 2012. Human serum samples were collected in 2007 and in 2011/2012 from a cohort of 2–29-year-old subjects from Mali, Senegal, and The Gambia. A sample that tested positive by Zika virus IgG ELISA and by Zika virus microneutralization test was defined as positive. In 2007, the highest prevalence was 21.9%, found in Senegal among 18–29-year-old subjects. In 2011/2012, the highest prevalence, 22.7%, was found still in Senegal, but in 11–17-year-old subjects. During both study periods, the lowest prevalence was found in Mali, where few positive cases were found only in 18–29-year-old subjects. The Gambia showed an intermediate prevalence. In the three countries, prevalence was strongly associated with increasing age. This study contributes to understanding Zika virus circulation within three different ecological and demographic contexts with scarce or no data currently available. Results showed that Zika virus circulated actively in West Africa between the period 2007 and 2011/2012, but with some geographic specificity.


Introduction
Zika virus (ZIKV) is a Flavivirus transmitted to humans mainly by Aedes mosquitoes that was first isolated in Uganda in 1947 [1,2]. Serological data suggest that ZIKV transmission has occurred among humans, animals and mosquitoes throughout tropical Africa for more than 70 years; however, ZIKV epidemics were never reported, and fewer than 20 human infections were recorded [3] between its isolation and the first large epidemic, occurring in Micronesia in 2007 [4]. ZIKV gained new attention after its spread in the Pacific and then to the Americas [5], and in recent years serological studies were conducted in Africa, documenting the presence of ZIKV antibodies in humans [6][7][8][9]. Moreover, a ZIKV outbreak that occurred in Gabon in 2007 was retrospectively identified [10]. Further, in October 2015, an outbreak of more than 7000 suspected cases was reported in Cape Verde with the identification of the

Materials and Methods
Human serum samples were collected during one of the clinical trials performed within the framework of the Meningitis Vaccine Project (MVP) [19]. The clinical trial was performed in 2007 in Mali, The Gambia, and Senegal to evaluate the safety and immunogenicity of MenAfriVac ® vaccine (Serum Institute of India, Pune, India) in comparison to a licensed comparator, Meningococcal Vaccine, in 2-29-year-old participants [20]. Acute disease (with or without fever) at the time of enrollment, administration of immunoglobulins and/or any blood products in the previous 30 days, administration of immunosuppressants or other immune-modifying agents in the previous 90 days, immunodeficiency or serious chronic illness, pregnancy or lactation were among the exclusion criteria. For each blood sample collected, a thick smear was examined for malaria parasitemia before vaccination. To evaluate antibody persistence to MenAfriVac ® vaccine (Serum Institute of India, Pune, India), blood samples from the same study participants were collected between 2011 and 2012 [21]. The study was conducted  under the International Council for Harmonization of Technical Requirements for Pharmaceuticals for  Human Use Good Clinical Practice (ICH-GCP) with identifier ISRCTN87739946. A total of 871 available samples were collected before vaccination between August and October 2007 from subjects of both sexes and aged 2-29 years in three different study sites: Bamako, Mali; Niakhar, Senegal; Basse Santa Su, The Gambia ( Figure 1).
Viruses 2020, 12, x FOR PEER REVIEW 3 of 9 [21]. The study was conducted under the International Council for Harmonization of Technical  Requirements for Pharmaceuticals for Human Use Good Clinical Practice (ICH-GCP) with identifier  ISRCTN87739946. A total of 871 available samples were collected before vaccination between August and October 2007 from subjects of both sexes and aged 2-29 years in three different study sites: Bamako, Mali; Niakhar, Senegal; Basse Santa Su, The Gambia ( Figure 1). For each study site, samples were stratified by age group (2-10, children; 11-17, adolescents; 18-29year-old, young adults) and sex, as shown in Table 1. All sera were tested for presence of ZIKV IgG antibodies by commercial ELISA (Euroimmun TM , Lübeck, Germany), following the manufacturers' instructions. Samples were considered positive when the ratio between the optical density (OD) of the sample and that of the calibrator was >1.1, negative when the ratio was <0.8.
All sera with borderline or detectable ZIKV IgG ELISA antibodies were tested by an in-house ZIKV microneutralization test (MNT) [22]. MNT was performed in a 96-well format based on cytopathic effect (CPE), using the ZIKV strain UVE/ZIKV/1947/UG/MR766 obtained from the For each study site, samples were stratified by age group (2-10, children; 11-17, adolescents; 18-29 year-old, young adults) and sex, as shown in Table 1. All sera were tested for presence of ZIKV IgG antibodies by commercial ELISA (Euroimmun TM , Lübeck, Germany), following the manufacturers' instructions. Samples were considered positive when the ratio between the optical density (OD) of the sample and that of the calibrator was >1.1, negative when the ratio was <0.8.
For the purpose of this study, we defined as positive to ZIKV a subject whose sample tested positive to ZIKV IgG ELISA and to ZIKV MNT. Statistical analyses were performed using SAS ® software version 9.4 (SAS Institute Inc., Cary, NC, USA). Sex-and age-specific prevalence rates and seroconversion rates of samples collected in 2007 and 2011/2012 from the same subjects were calculated for each study site along with the corresponding exact two-sided 95% confidence interval (CI) obtained using the Clopper-Pearson method. The association of prevalence rate with age groups and study sites during each study period was examined by a logistic regression model, adjusting for sex and malaria parasitaemia test result. The model for the later study period was additionally adjusted for the meningococcal group A (MenA) specific antibody titers measured by serum bactericidal antibody assay with rabbit complement (rSBA), 28 days following MenA vaccination, to explore any interaction between the immune response to MenAfriVac ® (Serum Institute of India, Pune, India) vaccination and immunity against ZIKV. Logistic analysis was also used to investigate the association of the seroconversion rate with age groups and study sites adjusting for sex. A significance level of 0.05 was used for the two-sided statistical tests.

Results
Out of 1521 samples, 142 showed borderline or detectable ZIKV IgG ELISA. A total of 95 samples out of 142 also tested positive to ZIKV-MNT and were considered positive for the purpose of this study.
When comparing serologic results between 2007 and 2011/2012 within the same subject, the highest seroconversion rate to ZIKV was observed in Senegal (4%, 95%CI 1.76-7.81) with seroconversion in six subjects aged between 11-17-years-old and one subject in each of the 2-10 and 18-29-year-old age groups, followed by The Gambia with seroconversion in six (2.9%, 95%CI 1.07-6.17) subjects, three in each 11-17-and 18-29-year-old age groups, and by Mali with only one subject (0.5%, 95%CI 0.01-2.5) seroconverted in the 18-29-year-old age group (Table 3). The differences in seroconversion rates among sites and among age groups were not statistically significant. No statistically significant differences were found between males and females either.  (Table 4) or with a ZIKV MNT titer < 1:40. In both study periods, differences in ZIKV prevalence rates between study sites were statistically significant (p < 0.0001 for 2007 and p = 0.001 for 2011/2012). No significant differences were found between males and females, while a clear trend of increasing ZIKV positivity was found with increasing age (p = 0.0004) (Table 2, Figure 2). No statistically significant association was found between ZIKV prevalence rate and malaria parasitaemia test result in either study period. The MenA rSBA titer measured at 28 days post-vaccination was not shown to be significantly associated with ZIKV prevalence rate in the 2011/2012 study period.
When comparing serologic results between 2007 and 2011/2012 within the same subject, the highest seroconversion rate to ZIKV was observed in Senegal (4%, 95%CI 1.76-7.81) with seroconversion in six subjects aged between 11-17-years-old and one subject in each of the 2-10 and 18-29-year-old age groups, followed by The Gambia with seroconversion in six (2.9%, 95%CI 1.07-6.17) subjects, three in each 11-17-and 18-29-year-old age groups, and by Mali with only one subject (0.5%, 95%CI 0.01-2.5) seroconverted in the 18-29-year-old age group (Table 3). The differences in seroconversion rates among sites and among age groups were not statistically significant. No statistically significant differences were found between males and females either.  (Table 4) or with a ZIKV MNT titer < 1:40. Differences or trends in increasing or decreasing MNT titer by age groups or sex were not statistically significant.

Discussion
The present study provides data on the circulation of ZIKV in West Africa during the years of its emergence in the Pacific region [4]. Results of this study indicate that in 2007 and 2011/2012, ZIKV was present and actively circulating in Senegal and The Gambia and yet virtually absent in Mali.
In Senegal, the overall prevalence of seropositivity to ZIKV was 13%, and was higher in subjects older than 11 years during both study periods. These findings confirm previous WHO reports, noting the presence of local cases of ZIKV infection in the region [16]. Samples were collected in Niakhar, a rural area that most likely allows exposure to both urban and sylvan mosquitoes. A study conducted in Senegal on samples collected from 1992 to 2016 [8] showed that ZIKV IgM prevalence among febrile patients was between 5% and 7.5%, demonstrating continuous transmission of ZIKV in Senegal over two decades. In our study, 4% of subjects from Senegal seroconverted for ZIKV between 2007 and 2011/2012, confirming the findings of Herrera et al. [8].
In our study, ZIKV prevalence in The Gambia was somewhat lower than in Senegal with no positive samples in children in both study periods. As for Senegal, samples from The Gambia were collected in a rural village with similar climatic and ecological conditions, including the presence of competent vectors, as in Niakhar. It is therefore difficult to explain the differences in ZIKV prevalence between the two areas. We could hypothesize that probably there is a difference in the Aedes population in The Gambia for local ecological reasons, or a heterogeneity due to the time and place of sampling. To our knowledge, this is the first report on the presence of ZIKV in humans from The Gambia, where the WHO report indicates only the presence of DENV and CHIKV as local cases [16].
Our findings show that ZIKV is virtually absent in Mali, with only a few positive subjects in the 18-29-year-old age group. One possible explanation is that, despite the urban environment being extremely favorable to the survival of Aedes aegypti in Bamako, transmission of ZIKV is reduced and almost absent compared to the rural study locations.
One interesting observation is that all the three countries included in this study implemented a similar malaria control program during the study period, leading to a steady decline in malaria cases up to a pre-elimination stage [24,25]. Many strategies erected against malaria and its vector, e.g., long-lasting insecticide-treated nets, are unlikely to significantly impact Aedes populations, because of biological and behavioral differences [26]. In addition, in our study no interaction between the presence of malaria parasitemia and ZIKV antibody presence was observed, suggesting that co-infections are common in sites where exposure to both Anopheles and Aedes vectors co-exist.
Our study confirms that ZIKV antibody acquisition is strongly related to increasing age. The increasing prevalence concurrent with age suggests that the population is consistently exposed to ZIKV throughout life. Furthermore, higher rates observed in the elderly may be the result of an earlier discrete exposure, followed by a decrease in ZIKV circulation that left younger cohorts unexposed.  [27]. However, in our study 52.8% of subjects had maintained the same ZIKV antibody titer in 2011/2012, suggesting that natural immunity against ZIKV may persist for at least five years. In 2016, during the epidemic reported in French Guyana, vertical transmission of ZIKV has been estimated to occur in 26% of foetuses of ZIKV-infected mothers. Among foetuses exposed to ZIKV through vertical transmission, 20% had severe complications suggestive of congenital Zika syndrome [28]. Results from studies conducted since the mid-1950s have suggested that ZIKV has likely been endemic for decades in sub-Saharan Africa [29], with a recently documented severe outbreak in Cape Verde, Guinea Bissau and Angola. To date, only one documented case of ZIKV congenital syndrome from the Asian lineage has been reported, identified in a child from Portugal, in which the virus was imported from Angola [13], where mosquito-borne ZIKV transmission has been reported [14] together with an increased incidence of congenital microcephaly [30]. The lack of reports on ZIKV congenital syndrome in Africa may be explained either by the immunity acquired by women prior to childbearing age following ZIKV infection that results in protection against congenital infection [29], or by the fact that birth defects caused by ZIKV might have gone unnoticed and are often attributed to other pathogens [31].
It is of interest that in 2007, the first period of our study, ZIKV was reported in Micronesia [4] from where it spread to other Pacific Islands [32,33]. Phylogenetic studies have shown that the ZIKV outbreak in Micronesia was initiated by a strain from Southeast Asia, revealing the existence of an Asian lineage of ZIKV, different from the African one [34,35]. As the ZIKV Asian lineage was introduced into Africa only in 2015 [1], we can assume that between 2007 and 2012, only ZIKV African lineage was circulating in our study sites.
Recent "in vitro" and non-human primate model studies demonstrated that immunity after infection with one ZIKV lineage provides cross-protection against another [36,37]. However, little is known regarding any differences between the African and Asian lineages in clinical manifestations as well as on vertical transmission and ZIKV congenital syndrome, although some findings suggest that the Asian lineage might cause less severe disease [38]. Given that both lineages are now circulating in Africa, there is a growing need for a better understanding of the current epidemiology and public health impact of ZIKV in Africa.