Investigation of a possible yellow fever epidemic and serosurvey for flavivirus infections in northern Cameroon, 1984

Bulletin of the World Health Organization, 65 (6): 855-860 (1987) Investigation of a possible yellow fever epidemic and serosurvey for flavivirus infections in northern Cameroon, 1984 T. F. TSAI, 1 J. S. LAZUICK,2 R. W. NGAH,3 P. C. MAFLAMBA,4 G. QUINCKE,5 & T. P. MONATH 6 A cluster offatal hepatitis cases in northern Cameroon in 1984 stimulated a field investigation to rule out an epidemic of yellow fever. A serosurvey of villages in the extreme north of the country, in a Sudan savanna (SS) phytogeographical zone, disclosed no evidence ofrecent yellowfever infection. However, further south, in a Guinea savanna (GS) phytogeographical zone, serological evidence was found of endemic yellow fever virus transmission. The results indicate a potential for epidemic spread of yellow fever virus from the southern GS zone to the nothern SS zone of Cameroon, where immunity in the population was low. In April 1984, the Ministry of Health of Cameroon and the WHO Regional Office for Africa, Brazza- ville, requested the assistance of the Centers for Disease Control in ascertaining whether yellow fever was the cause of a cluster of fatal cases of hepatitis in the northern province of the country and whether a larger outbreak had occurred there. The index case was a pregnant woman who died of hepatitis in January 1984 in Maroua, a city in the sub-Sahelian northern province (see Fig. 1). Two more fatal cases of hepatitis among pregnant women were reported in January 1984 in the same city. A histopathological diagnosis of yellow fever was made in the index case, but in no instance was a specific diagnosis of yellow fever confirmed. Entomological surveys in 1981 and 1983 found that, during the rainy season, the extreme l Medical Officer, Division of Vector-Borne Viral Diseases, Centers for Disease Control, P.O. Box 2087, Fort Collins, CO 80522-2087, USA. Requests for reprints should be sent to this address. 2 Research Biologist, Division of Vector-Borne Viral Diseases, Centers for Disease Control, Fort Collins, CO, USA. 3 Specialist in Public Health, Ministry of Public Health, Yaounde, Cameroon. 4 Director of Preventive Medicine, Ministry of Public Health, Yaound,6, Cameroon. s WHO Programme Coordinator, World Health Organization, Yaounde, Cameroon. At present: Responsible Officer, Food Aid Programme, WHO, Geneva, Switzerland. 6 Director, Division of Vector-Borne Viral Diseases, Centers for Disease Control, Fort Collins, CO, USA. north of Cameroon was extensively infested with Aedes aegypti as well as with the sylvatic vectors A. furcifer-taylori and A. vittatus (M. Germain, unpublished report, 1981; R. Cordellier, unpublished report, 1983). However, these surveys and a survey in Maroua in 1984, which included the dwellings of two of the fatal hepatitis cases, found no evidence for A. aegypti in the northern province during the dry season (J. P. Adams, unpublished report, 1984). We report here the results of our investigation to determine whether an outbreak of yellow fever had recently occurred in the north of Cameroon and whether conditions there were suitable for main- tenance of yellow fever virus with the potential for epidemic spread to urban centres. METHODS It was not possible to trace and confirm the diagnosis of yellow fever in surviving patients; therefore, our efforts focused on serological surveys. A mass immunization campaign against yellow fever had last been performed in Cameroon in 1971. Con- sequently, the serosurveys focused on children aged of 13 years or less, who would have been born after the campaign. Samples of blood were taken from volunteers in five villages near Maroua in the north- ern, dry (Sudan) savanna (SS) phytogeographical 4840 -855- T. F. TSAI ET AL. zone of the country (Fig. 1). To establish whether sylvatic transmission of yellow fever might be pre- valent among monkeys in the area, we also collected blood from Cercopithecus aethiops and Erythrocebus patas monkeys caught near Kossa. Further sero- surveys were conducted in villages approximately 250 km further south (Poli, Fignole, and Gode), which lie in the moist (Guinea) savanna (GS) phyto- geographical zone, corresponding to emergent zones of yellow fever elsewhere in West Africa (1). Although the disease had not been reported from this area, the moist savanna is capable of supporting high densities of vectors, suggesting that yellow fever might be enzootic and endemic in monkeys and humans, respectively. Sera were tested by IgM capture enzyme immuno- assay, haemagglutination inhibition (HI), comple- ment-fixation (CF), and neutralization tests for antibody to yellow fever and selected other flavi- viruses (Table 1) (2, 3). Approximately 300 sera were tested in the field by enzyme immunoassay. None of the sera had IgM antibody to any of the flaviviruses tested, which was not unexpected, since the half-life of IgM antibody is short and sera were collected nearly 6 months after the rainy season, Fig. 1. Northern region of Cameroon, showin of yellow fever investigation, 1984. Table 1. Schedule for the serological survey in Cameroon, 1984 1.All sera were screened by enzyme immunoassay for IgM antibodies to the following viruses:/bellow fever, West Nile, and either Murray Valley encephalitis or Koutango. 2.All sera were screened by haemagglutination-inhibition (HI) for antibodies to yellow fever, West Nile, and Zika viruses. Sera with HI titres > 1:80 to any antigen were: -tested further by complement fixation (CF) for antibodies to yellow fever, West Nile, Zika, Sepik, and Uganda S viruses, as well as for antibody to dengue virus type 2; and -screened for neutralizing antibody to yellow fever virus by plaque reduction neutralization (90% reduction was taken as the endpoint) in Vero cells. Sera with yellow fever virus neutralizing titres > 1:16 were further tested for neutral- izing antibodies to dengue virus type 2 as well as to Bouboui, West Nile, Zika, Sepik, and Uganda S viruses (selected sera) to determine specificity. when the most recent infections would have occurred. The interpretation of the HI, CF, and neutralization antibody responses was based upon a previously published scheme developed to differentiate primary from flavivirus superinfections (4). Type-specific re- sponses were classified as either monotypic, if antibody to a single antigen was detected, or homo- typic, if heterologous antibody titres were four times lower than the homologous response. RESULTS The results of the HI screening test showed that flavivirus infections were more prevalent in the northern SS zone than in the southern GS zone. CHAD Approximately 70-80% ofpersons in the SS zone had antibodies to yellow fever, West Nile, or Zika :OC Isotherm viruses, while in 70-80% of persons in the GS zone no antibodies to these antigens were detected in the HI test (Table 2). On first appearances, the prevalence of yellow fever antibody seems high in the northern zone. However, stratification of the results indicates that only 9% of the participants in the north with positive seroresponses were monotypic, while a significantly greater proportion (34%) of responses to yellow fever virus in the south were monotypic (Table 3). These 24*C observations indicate that flavivirus superinfections P were responsible for the apparent high prevalence of antibodies to yellow fever virus detected in the HI test vai,, in the north but that primary infection with the virus was more prevalent in the south. g the sites The results of CF tests confirmed these indications(Table 4). Again, taking monotypic responses as 856 FLAVIVIRUS SEROSURVEY AND POSSIBLE YELLOW FEVER IN NORTHERN CAMEROON 857 Table 2. Distribution of reciprocal titres in the haemagglutination-inhibition (HI) test for selected flavivirus antibodies by phytogeographical zone, Cameroon, 1984 No. of sera Yellow fever virus West Nile virus Zika virus Reciprocal Hi titre SS zone' GS zoneb SS zone GS zone SS zone GS zone <10 94 (16)c 182 (72) 141 (24) 201 (79) 204 (35) 209 (82) 10 85 (14) 21 (8) 109 (19) 34 (13) 144 (25) 23 (9) 20 72 (12) 12 (5) 105 (18) 10 (4) 69 (12) 8 (3) 40 59 (10) 13 (5) 89 (15) 5 (2) 47 (8) 6 (2) >80 276 (47) 26 (10) 143 (24) 4 (2) 122 (21) 8 (3) Total 586 (99) 254 (100) 586 (100) 254 (100) 586 (101) 254 (99) a SS = Sudan savanna. b GS -Guinea savanna. c Figures in parentheses are percentages. specific indication of infection with yellow fever virus, we found little evidence that the virus had recently been active in the SS zone; however, recent infection may have occurred in as many as 2.4% of the population of the GS zone. In contrast, West Nile virus infections were prevalent in the northern zone, where more than 9% of the population sample had evidence of recent infection with the virus (monotypic CF antibody). The results of the neutralization tests confirmed the paucity of yellow fever activity in the northern SS zone (Table 5). Of 90 sera tested against yellow fever Table 3. Response pattern in the haemagglutination-inhibition (HI) antibody test for selected flaviviruses by phyto- geographical zone, Cameroon, 1984 No. of sera' Zone Yellow fever virus West Nile virus Zika virus Sudan savanna 43/492 (9) 16/446 (4) 0/382 Guinea savanna 24/70 (34) 12/53 (23) 6/45 (13) a Shown is the monotypic HI response/all HI responses. Figures in parentheses are percentages. Table 4. Response pattern in the complement-fixation (CF) antibody test for selected flaviviruses by phyto- geographical zone, Cameroon, 1984 No. of sera' Zone Yellow fever virus West Nile virus Sepik virus DEN-2b Uganda S virus Zika virus Sudan savanna 4/50 (0.7) 55/129 (9.4) 2/51 10/105 5/75 0/37 Guinea savanna 6/11 (2.4) 4/6 (1.6) 1/5 2/4 1/3 0/2 a Shown is monotypic CF response/all CF responses. Figures in parentheses are the estimated point prevalences of monotypic antibody in the population sample: Sudan savanna zone, n = 586; Guinea savanna zone, n = 254. b DEN-2 = dengue virus type 2. T. F. TSAI ET AL. Table 5. Distribution of neutralizing antibodies for selected flaviviruses by phytogeographical zone, Cameroon, 1984 No. of seraa Zone Yellow fever virus West Nile virus Bouboui virus Zika virus DEN-2/Sepik virusb Indeterminate Total Sudan savanna 10 (1.7) 47 (8.0) 0 1 0 32 90 Guinea savanna 17 (6.7) 2 (0.8) 2 0 0 3 24 a Shown are monotypic and homotypic responses. Figures in parentheses are the estimated point prevalences of specific antibody in the population sample: Sudan savanna zone, n - 586; Guinea savanna zone, n -254. b DEN-2 = dengue virus type 2. and other flaviviruses, homotypic responses to the yellow fever virus were found in only 10 persons from the SS zone-all of whom were either over 14 years of age, who may have been immunized, or students at a secondary school, whose ages may not have been reliably given. In the GS zone, 17 persons had type-specific antibody to yellow fever virus. The median age of those who had positive seroresponses in the latter zone was 9 years, while in the north it was 12 years. The point prevalence of yellow fever viral antibody in the SS zone (1.7%) was significantly lower than that in the GS zone (6.7%) (P<0.001, Poisson). The serosurvey did not aim to establish the pre- valence of West Nile viral infections (sera were selected for neutralization tests on the basis of yellow fever virus neutralizing antibody titres of > 1:16, Table 1). However, of the sera tested, 47 from the SS zone had homotypic neutralizing antibody to West Nile virus, which is equivalent to an estimated point prevalence of 8 %, and is significantly higher than the estimated prevalence in the GS zone (P <0.001, Poisson). The serological survey of monkeys caught near Kossa established that sylvatic yellow fever did not occur in the SS zone and confirmed that West Nile Table 6. Reciprocal titres of neutralizing antibodies for selected flaviviruses in monkeys, Kossa (Sudan savanna zone), Cameroon, 1984 Reciprocal titre Reference Yellow fever West Nile Zika Uganda S DEN-2a Bouboui Species number virus virus virus virus virus virus Cercopithecus aethiops { 75136 16 128 32 <4 <4 NAb75139 4 64 <4 NA 64 32 75140 4 256 <4 NA 32 16 75141 4 256 <4 NA <16 16 75142 4 128 <4 NA < 16 < 16 75143 4 64 <4 NA 32 16 75144 <4 NA NA NA NA NA 75145 4 256 <4 NA 64 16 Erythrocebus patas 75146 8 256 <4 NA 64 16 75147 < 4 NA NA NA NA NA 75148 4 256 <4 NA 64 16 75149 32 256 <4 <4 <4 NA 75150 4 128 <4 NA 16 < 16 75151 8 64 <4 NA 32 <16 75152 8 256 <4 NA 32 < 16 DEN-2 = dengue virus type 2. b NA - Not available. 858 FLAVIVIRUS SEROSURVEY AND POSSIBLE YELLOW FEVER IN NORTHERN CAMEROON virus was the predominant flavivirus infection in the north of Cameroon (Table 6). None of the 15 monkeys tested had evidence of infection with yellow fever virus, and all 11 of the 13 monkeys that were tested for antibody to other flaviviruses had homo- typic responses to West Nile virus. DISCUSSION The serological surveys reported indicate that yellow fever was not as widespread in the north of Cameroon as would have been expected had an outbreak recently occurred there. Further evidence against the occurrence of yellow fever in this part of the country came from histopathological reviews of the liver sections from the index case carried out independently by three consultants, one each at the Institut Pasteur, Paris; the Armed Forces Institute of Pathology, Washington, DC; and the Centers for Disease Control, Atlanta, GA. The consultants con- cluded unanimously that massive hepatic necrosis, consistent with viral hepatitis, was present. However, consultants A and B pointed out that microvesicular changes in the hepatocytes, the absence of Council- man's bodies, and the presence of an inflammatory infiltrate were features of the case that were atypical of yellow fever and indicated another viral etiology. Consultant C reserved judgment but agreed that the histopathology was not diagnostic of yellow fever. The limitations of diagnosing yellow fever based on histopathological findings alone are underscored by the discrepancies in the experts' diagnoses with the diagnosis of yellow fever made earlier in Yaounde, which led to the investigation. The combination of serological data, the results of entomological surveys, and the consultants' histo- pathological diagnoses indicate that yellow fever was not endemic in the northern province of Cameroon in early 1984. The serosurvey of monkeys confirmed the absence of sylvatic yellow fever in the SS zone. However, in the more southerly GS zone, sero- surveys disclosed endemic transmission of yellow fever virus, as suspected. Therefore, conditions exist in Cameroon for maintenance of yellow fever virus in endemic areas, with the possibility of its emergence and introduction to susceptible populations in the north, resulting in subsequent epidemic spread. We recommend improvements in the existing sur- veillance programmes for yellow fever in Cameroon and establishment of local laboratory facilities that would facilitate a timely, specific diagnosis of the disease in the country. The continuation of limited immunization campaigns in the north of the country and the inclusion of yellow fever immunization in the expanded programme on immunization are control measures being considered by the Ministry of Public Health. ACKNOWLEDGEMENTS We thank W. Veseley and D. Muth for technical assistance. The investigation was sponsored by WHO. RESUME ETUDE SUR UNE tVENTUELLE EPIDEMIE DE FIEVRE JAUNE ET ENQUETE StROLOGIQUE SUR LES INFECTIONS A FLAVIVIRUS DANS LE NORD DU CAMEROUN, 1984 En 1984, plusieurs cas mortels d'hepatite survenus dans le nord du Cameroun ont conduit a entreprendre une etude de terrain afin d'ecarter l'hypothese d'une epidemie de fievre jaune. Des enquetes serologiques ont ete realisdes dans des villages de l'extreme nord du pays (zone phyto- geographique de savane soudanaise) et dans des villages situes 250 km plus au sud (zone phytogeographique de savane guineenne). L'enquete a ete axee sur les enfants de moins de 13 ans, nes apres la derniere campagne nationale de vaccination de masse contre la fievre jaune. Elle n'a montr6 aucun signe de transmission recente de la fievre jaune dans la zone de savane soudanaise, ce qui a permis d'exclure l'hypothese d'une 6pidemie. Dans cette zone, moins de 2% de la population presentaient des anticorps neutralisants contre la fievre jaune, et 15 singes captures aux alentours des villages d'etude n'ont montre aucun signe serologique d'infection par le virus de la fievre jaune, ce qui semble confirmer l'absence de transmission endemique et en- zootique de la fievre jaune dans cette zone. Toutefois, 859 860 T. F. TSAI ET AL. dans 1'echantillon de population observe, 8% des enfants pr6sentaient des anticorps diriges contre le virus West Nile. Dans le sud, c'est-a-dire en zone de savane guineenne, on a trouvd des signes d'end6micite de la fievre jaune: des infections recentes, confirmees par la presence d'anticorps fixant le compldment, ont etd decelees chez 2,4% des enfants et 6,7% d'entre eux possedaient des anticorps neutralisants contre le virus de la fievre jaune. Les resultats de ces etudes montrent qu'il n'y a pas eu de transmission recente du virus de la fievre jaune dans le nord, c'est-a-dire en zone de savane soudanaise, et que cette region ne connait pas de transmission endemique du virus. La population de cette zone est donc sensible a l'infection et une epidemie se declencherait si le virus etait introduit. En revanche, la transmission du virus de la fievre jaune est endemique dans le sud, en zone de savane guineenne. I1 existe ainsi un foyer d'activite permanente de la fievre jaune, a partir duquel le virus peut se propager, a 250 km seulement des populations urbaines sensibles du nord. REFERENCES 1. GERMAIN, M. ET AL. Recent advances in research regarding sylvatic yellow fever in West Central Africa. Bulletin de l 'Institut Pasteur, 80: 315-330 (1982). 2. SHOPE, R. E. & SATHER, G. E. Arboviruses. In: Lennette, E. H. & Schmidt, N. S., ed. Diagnostic procedures for viral, rickettsial, and chlamydial in- fections. Washington, DC, American Public Health Association, 1979. 3. SALUZZO, J. F. ET AL. Interet du titrage par ELISA des IgM sp6cifiques pour le diagnostic et la surveillance de la circulation selvatique des flavivirus en Afrique. Annales de l 'Institut Pasteur, 137: 155-170 (1986). 4. MONATH, T. P. ET AL. The 1970 yellow fever epidemic in Okwoga District: Benue Plateau State, Nigeria. 3. Serological responses in persons with and without any pre-existing heterologous Group B immunity. Bulletin ofthe World Health Organization, 49: 235-244 (1973).