Onchocerciasis infection in children born during 14 years of Simulium control in West Africa

TRANSACTIONS OF THE ROYAL SOCIETYOF TROPICALMEDICINEAND HYGIENE(1991), 85, 385-390

Onchocerciasis infection control in West Africa G. De Sole and J. Remme Burkina Faso

in children

Onchocerciasis

Control

Abstract

The incidenceof onchocerciasisinfection in children born sincethe start of vector control is oneof the indicators usedin the epidemiologicalevaluation of the OnchocerciasisControl Programme in West Africa (OCP). Though initially of little value, after a decadeof control it hasbecomea sensitiveindicator of residual transmission.The results of 14 years of control are reported. In 179 villages parasitological surveys were undertaken at intervals of 3-l years. 15 286children wereexaminedand 110werefound to be infected, comparedto an expectednumberof 2467 infected had there been no control. There was considerablegeographicalvariation in the results. In the large central OCP areathe resultswere excellent. Of 12 172children examinedin 127villages,only 23 were found to be infected comparedto & expected number of 1960without control. This suggeststhat larviciding had achieved a 99% reduction in the incidenceof infection in children. Additional surveys in 2 foci in the central OCP areawhere transmission had relapsedshowedthat theseproblemswere very localized. Most villages with infected children were found in OCP border-areas in the eastandwest, which had been reinvaded by infective vectors from elsewhere, and in the intermediatearea betweenforest and savannain C&e d’Ivoire where there had been partial control failures becauseof resistance.The incidenceof infection in children wasreducedby an estimated68% in the easternreinvadedarea,by 87% in the western reinvaded area, and by 84% in the intermediate area. Introduction

Since1975,the OnchocerciasisControl Programme in West Africa (OCP) has attemnted to control the vector of onchocerciasis,Simuliuh damnosum s.l., by aerialapplication of larvicides to the vector breeding sites. The aim of this strategy was to interrupt onchocerciasistransmissionfor a sufficiently long period to allow the reservoir of the parasite, Onchocerca volvulus, to decreaseto such a low level that larviciding could be interrupted without having a seriousrisk of recrudescenceof the disease(WHO, 1987a).The required duration of successfulvector control has recently been estimatedto be 14 years (PLANER et al., in press). Throughout the control neriod narasitologicalsurveys have been undertaken at r&&u int&vals in selectedindicator villagesfrom all major river basins of the Programmearea. The effectivenessof vector control hasbeenevaluatedby comparingthe observed trends in the prevalenceand intensity of infection in theseindicator villageswith the trends predicted on the assumptionof no interruption of transmission iF&O, 1987b; DE SOLE et al., 1990;REMME et al.,

born

during

Programme

385

14 years

in West Africa,

of Simulium

B.P.549,

Ouagadougou,

Another index which has been used in the epidemiological evaluation is the incidence of onchocerciasisinfection in children born since the start of control (BA et al., 1987).Becauseof intrinsic limitations of this index, it has been of little value during the first 5-8 years of contol (REMME et al., 1986): However, during later years it has given important confirmation of the effectivenessof vector control in most of the central OCP areawhile it has also provided unambiguous evidence of residual transmissionin certain problem foci. Additional studiesof the prevalenceof infection in children born sincethe start of control in non-indicatorvillageshave enabledthe geographicaldelimitation of theseproblem foci. In-&& paper we report on onchocerciasis infectionsin children born durina the first 14vearsof vector control in the OCP. Materials and Methods Selection of indicator villages

The indicator villageswere selectedto renresentall river basins in the Programme area where onchocerciasiswas endemic. Most of the villages chosen were hyperendemic with a prevalence of infection in excessof 60%. Becauseof the long follow-un neriod foreseen.the indicator vihaaeshad to have a siablepopulationwith little or no n&ration and the village sizehad to be around 300inhabitants. Between 1975and 1979, 338 villageswere surveyed for the first time. In the presentpaper we report the resultsfor 179of those villageswhich were retained for follow-up surveysat a 3- to 4-yearsinterval, and which had their last survey after 1985. The cut-off point of 1985wasadoptedbecausethe last published accountof infectionsin children in the OCP covered the period up to 1984inclusive (BA et al., 1987).In 147 of the 179villages, the last survey was done in 1989or 1990after 12 to 14 years of vector control. Routine

skin snip surveys

A routine survey consistedof a completecensusof the village population and a parasitologicalexamination. In the latter a skin snipwastaken from eachiliac crest of the villagers, using a Holth sclero-cornea1 punch, and the snipswereincubatedin distilledwater for 30 min. Eachskin snip wasthen examinedat low magnificationfor the presenceof 0. volvulus microfilariae and the number per snip counted (PROST & PROD'HON, 1978). All negativesnipswere incubated for a further 24 h in saline and subsequentlyreexamined. The results were recorded on standard OCP forms. As the censusclerk filled the standardform during the censushe noted on a separatesheetof paper the examinationcode, name,age,sexand family number of every child whoseagewasequalto or lessthan the number of years the village had been under vector

386 control. At the end of the day’s work the census clerk checked the results recorded by the microscopist against the list of children. If any child was found to be infected, skin snips were taken again from this child and examined using the standard procedure. This verification was needed to avoid errors due to mis-labelling. If the infection of a child was thus confirmed, a migration history was taken to help to assess whether the infection had been contracted locally or not. Routine skin snip surveys were also undertaken in 3 villages from a non-controlled onchocerciasis area in the southern extension of the OCP in 1979, 1983 and 1987, but vector control was not started in the southern extension until 1988. These 3 villages have been included in the analysis for the purpose of comparison between controlled and non-controlled areas. Detailed

surveys in non-indicator

villages

In 2 foci in the central OCP area, where the epidemiologicalevaluationhaddemonstrateda partial failure of vector control! detailed surveys were undertakenin severalnon-mdicatorvillagesto determine the geographical limits of the area where transmissionhad occurred. These detailed surveys involved only examinationof children who had been born sincethe start of control. For children below the ageof 10 yearsthe dental formula wasusedto aid age estimation.The parasitologicalexaminationwasdone Ratio

accordingto the standardproceduredescribedabove. Data analysis

The skin snipdata wereprocessed and analysedon microcomputersusingcomputerprogramsdeveloped in the OCP. In the analysisfor the indicator villagesa child was defined as being born since the start of control if it hadnot beenregisteredduring the census of the baselinesurvey, which wasusuallydoneduring the first or secondyear of control, but if it was registeredduring the censusof one of the follow-up surveysasbeing born sincethe last survey. Sinceno longitudinal datawere availablefor the non-indicator villages,during the detailedsurveysthe identification of children born sincethe start of control wasbased on the estimatedage of the child only. The observednumber of infected children in the indicator villages was compared with the number expectedhad there beenno vector control. For each child born sincethe start of control the probability that, without control, it would have beeninfected by the ageit had reachedwhen it waslast examinedwas estimatedfrom the proportion of children of that year of agewho wereinfectedduring the baselinesurvey in the samevillage. The probabilitieswereaddedfor all children born sincethe start of control who hadhadat leastonefollow-up examination,and the sumgavethe expected number of infected children in the village who would have beendetectedduring the parasitological surveys if there had been no control. For the

between

,I

CENTRAL

WESTERN REINVADED! AREA

.’

e-p onyQ~~-.x-

t

‘,pJL

-----

-Q

i-./@

d

I

0

c:: 50

Programme boundary Area wth detoded survey 1001;m

Fig. 1. Location of indicator villages in the central Onchocerciasis Control Programme area and the ratio between expected number of infected children born since the start of vector control in each village.

observed

and

387

In the 179 indicator villages 15286 children born sincethe start of control wereexaminedat leastonce. The total numberof children found infected was110, compared to an expected number of 2467 without control. This suggests that vector control achievedan overall reduction-of about 95% in the incidenceof infection in children. However, there wasconsiderable geographicalvariation in the results. This is shown in Fie. 1. which eives the location of each village in the”Programme&ea and the ratio between the observedand expected number of infected children. Most of the villageswith infected children were in the east,westand south-westof the original Programmearea.This wasconsistentwith previousentomological findings which demonstratedresidualtransmissionin theseareasasa resultof reinvasionby infected vectors from outside the Programmearea and larvitide resistancein S. sanctipauli s.1.in the intermediate zone in C&e d’Ivoire (WALSHet al., 1979; WHO, 1987a;PHILIP~ONet al., 1990).In the Table, therefore, the results are presentedseparatelyfor four areas,i.e. the non-reinvadedcentral OCP area, the western reinvaded area, the easternreinvaded area and the intermediate area in C&e d’Ivoire. Central OCP area. In the central OCP area 23 children were found to be infected, 1.2% of the number expected without vector control. In more than 90% of the 122 villages surveyed no infected children were found. In the villages with infected children the ratio between observedand expected number of infected children was generally lessthan 0.1. The only exceptionswere the village of Pendit along the River Dienkoa in Burkina Faso and the village of Gorebaalong the River Kulda in Ghana, which had observed/expectedratios of O-39and 0.20 respectively. These 2 villages were located in 2 problem foci where detailed surveys were done, the resultsof which are reported below. Theseproblem foci accountedfor 10 infected children. Of the remaining 13 infected children from the central OCP area.6 hadmigratedfor oeriodsof 1 to 5 years to endemic onchoc&ciasiszones outside the

Programme area; it was not possible, therefore, to determine whether they had contracted their infection locally or in a non-controlled area. The other 7 cases had no history of migration and represented infections due to local onchocerciasis transmission. These included 2 cases from the middle Como6 river in C&e d’Ivoire, 1 case from an affluent of the White Bandama river in CBte d’Ivoire, 2 cases from the upper Comae in western Burkina Faso, 1 case from the River Sissili in Ghana, and 1 case from the River Koulpeolgo in eastern Burkina Faso. The latter was a very special case, an infected child in a village which was previously hyperendemic but where all other members of the population had become skin snip negative during the last survey. Western and eastern reinvasion areas. The results were lessfavourable for the reinvaded areas.In the eastern reinvaded area 30 infected children were detected, 32% of the number expected. The worst resultswereobtainedfor 3 villagesalongthe Kara and MB rivers, where23 childrenwere infected compared with an expectednumberof 33. No infected children were found in 4 villages in the reinvaded area of Benin. However. in two of thesevillaaesthe exwcted numbers of k&ted children were-zero and 0.09 respectively. In the western reinvaded area 35 children were detected, 12.7% of the number expected without control. Of these. 12 children camefrom 5 villages located along affiuents of the River Niger in yhe northern half of the westernreinvadedarea, and 23 infected children from 4 villagesin the southernhalf located along affluents of thg Rivers Sassandraand Marahoue. Two lamevillagesalonethe River Sassandra in C&e d’Ivo;e acc&mted f& as many as 20 infected children, but this wasstill lessthan 25% of the number exnected without control. In relative termsthe worst resultswereobtainedfor the 6 villages surveved alone the River Baed in Mali. where 10 infec&d cbil&en were obse&edin 4 vfijages compared with an expected number of 21.5 for the 6 villages together. Zntmdiate area. In the intermediateareain C&e d’Ivoire the resultswere highly variable. No infected children were detected in the western Dart of the intermediateareabut 22 infected children-werefound in the easternhalf. Of these, 15 camefrom the most southern village in Fig. 1, on the border of the Programmearea near the conlluenceof the Rivers

Table.

born since the start of onchocetciasis

non-indicatorvillagesit wasnot possibleto calculate the expectednumber of infected children becauseof the lack of baselinedata. Results Routine skin snip surveys

Results

of skin snip examinations

Vector control area Central OCP” Western reinvaded Eastern reinvaded Intermediate Total vector control area Non-controlled area

of children

Number of villages surveyed 127 22

Number of villages with infected children 12 9

:; 179

i 34

3

3

“OnchocerciasisControl Programme.

Number of children examined 12172 1230 810 1074 15286 146

Number of infected children Observed Expected 23 1959.9 % 275.6 93.9 22 137.8 110 2467.2 22

27.2

control

Ratio of observed to expected o-012 O-127 O-320 0.160 o-045 0.809

388 Bandama and Marahoue. The other 7 children came from 5 different villages along the Rivers Bandama, N’zi and Comot. It should be noted that the result showed in Fig. 1 for the village with an observed/ expected ratio of 0.71 was inflated, as it represented a single infected child against an expected number of 1.4 only. Non-controlled area. The Table also shows the results for the 3 villages in south Benin, where standardskin snip surveyswere doneduring a period of 8 years without vector control. Of 146 children born in thesevillagessincethe first survey in 1979,22 were found to be infected during the follow-up surveys. The difference betweenthis observednumber of infected children and the expectednumber of 27.2 was not statistically significant (binomial distribution, P=O*16). Detailed

surveys in the central

OCP area

Dienkoafocus. Cross-sectional surveyswere undertaken in 17 villages from the Dienkoa focus in Burkina Faso; 1383children born sincethe start of control in 1975were examinedand 26 of them were found to be infected. Theseinfected children came from 6 villages, including the indicator village of Pendiementionedearlier. The locationof the villages and the prevalenceof infection in children born since the start of control are shownin Fig. 2. The 6 villages with infected children were all close to the major

Fig. 2. Location of surveyed village in the Dienkoa children born since the start of control in each

vector breeding sitesalong a 15 km stretch of river between the village of Fon Hameau and Pendie proper, but no infected children were found in the surroundingvillagesat a distanceof morethan 5 km from this river stretch. KulpawnlKulda focus. Cross-sectional surveyswere alsodonearoundthe indicator village of Gorebaalong the River Kulda, an affluent of the White Volta in Ghana, where 4 infected children had been detected during the routine surveys. Additional surveyswere done in villages along the nearby Kulpawn river, where an infected child was detectedin eachof the indicator villages of Kulun and Yagaba. For all 3 indicator villagesit had beenshown previously that the declinein community microlilarial load (CMFL; .REMME et al., 1986)wasunsatisfactoryand there was epidemiologicalevidence of residual transmission (WHO, 1987b). Surveysweredone in 24 villagesand 1642children born sincethe start of control wereexamined(seeFig. 3). Of these,20 children in 6 villageswerefound to be infected, including the 6 infected children previously identified in the 3 indicator villagesof Goreba,Kulun and Yagaba.Around Gorebathe problemappearedto be very limited in spaceand no infectedchildren were found in any of the surrounding villages, with the exceptionof the village of Weni whereone child was infected. However, this child wasan immigrant from an endemicareasouth of the Programmeboundary

focus in Burkina Faso and the prevalence of infection village (the location of the focus is shown in Fig.

in 1).

389

Fig. 3. Location of surveyedvillagesin the Kulpawn/Ktdada focusin Ghanaand the prevalenceof infection in children born since the start of control in each village (the location of the focus is shown in Fig. 1).

and most probably had contracted the infection elsewhere. Along the Kulpawn river there were infected children in all 4 villagessurveyed; the worst affected wasBlielikpong. Not only were 12 infected children found. but the nrevalenceof infecton in the total populationwasstill 54.3%and the CMFL wasashigh as 18.2 microfilariae per snip. No infected children were detected in the 4 villages surveyed to the north-east of the Kulpawn valley. Discussion

Sincethe start of the epidemiologicalevaluationin the OCP it was evident that incidence of infection could not be adequatelymeasured.Any re-infection of infected individuals could not be detected by the skin sniptechniqueand the majority of the population of the indicator villageswasalready infected. Furthermore,even thosewho wereskin snipnegativeat the baselinesurvey and positive at a follow-up survev would not necessarilyrepresentincident casesbecause falselv negative classificationsin liehtlv infected individualsarenot uncommon(DE SOLE eial., 1990). Infection in children born sincethe start of control could be the only specificindicator of new infections occurring during the vector control period. The sensitivitv of this indicator denendsgreatlv on the duration of *vector control. Dur$g th; first-few years of control the sensitivity is very low because onchocercalinfections are rare in young children. However, as the period of control gets longer, the number of children born since the start of control increases and, most importantly, the older onesmove into agegrottoswhich would be more exoosedto the vectoFin;he event of a breakdownof co&o1 (REMME et al., 1986). This explains why, since the previous report on infections in children, which covered the first 10 yearsof control in the OCP (BA et al., 1987), the number of children born sincethe start of control has increasedby 63% from 52 to 85 children per village, while the expectednumber of infected child-

ren increasedby asmuch as228%, from 4.2 to 13.8 infected children per village. The sensitivity of infections in children asan indicator of the incidence of infection, therefore, has considerably increased during the later years of control. The resultsof the analysisconfirm the impressive successof vector control in the OCP. In the total original OCP area, the incidence of infection in children was reduced by 95%, and the results were particularly gratifying for the large non-reinvaded central OCP area. where 99% reduction in incidence wasachieved.With the exceptionof 2 known problem foci, the few infections detected in the central area were either imported or isolated casesof local infection. In more than 90%of the villagesnot a single new infection was detected, and the results with children seemto confirm that onchocerciasistransmissionhas been virtually interrupted over most of the central OCP area(WHO, 1987b;DE SOLE et al., 1990; REMME et al.,

1990).

Interruption of transmissionwas not achieved in the reinvaded zonesin the west and the eastof the OCP, even though it wasestimatedthat the incidence of infection in children was reduced bv 68% in the easternreinvaded area and by 87% in the western reinvadedarea.The continued but reducedtransmission in the westernand easternreinvaded areashas been clearly documentedby different entomological and epidemiologicalindicators (WHO, 1987b; DE SOLE et al., 1990; PHILIPPON et al., 1990). All these indicatorsdescribedthe samepattern of infectionsin children and showedthat transmissionhad occurred throughout the reinvaded areas but with varying degreesof intensity. It is only recently that this problem has been attacked fundamentally with the extensionof vector control activities to the sourcesof the reinvasion (PHILIPPON et al., 1990). In the intermediatezone in Cote d’Ivoire there was an 84% reduction in infections in children, or 92% reduction if one village on the boundary of the Programmearea was not taken into account. These

390

results are better than expected given the problems with larvicide resistance and the high biting rates of S . sanctipauli s.1. which have been observed in different ;ars in parts of the intermediate area (PI-IILIPPON et 1990). Nevertheless, there clearly has been rezdual transmission during the control period, and this has consequences for long-term control in this area. However, it was not everywhere clear whether the transmitted parasite strain was the severe blinding savanna strain o-r the more benign forest strain of 0: volvulus (DADZIE et al.. 1990). The promisine results which have been obtained recently with a de&ribonucleic acid probe for differentiating forest and savanna strains of 0. volvulus are therefore of great practical interest for control in the intermediate area (ERTTMANN et al., 1990). Detailed study of the 2 foci in the central area where transmis&on had recurred has shown the problem to be very localized, notably in the Dienkoa focus and around the village of Goreba along the River Kulda. Nearlv all infected children came from first-line villa& close to the breeding sites, and this underlines once more the importance of using firstline villages as indicator villages for the epidemiological evahation. The fact that the problem foci were very small was, of course, fortunate. However, it raises also some concern that other, equally small, foci with residual transmission may have remained undetected by the routine entomological and epidemiological evaluation. It is partly for this reason that detailed epidemiological mapping is undertaken of any area in the OCP for which cessation of larviciding is being considered (DE SOLE & REMME, in press). Acknowledgements We thank Dr E. M. Samba, Director of OCP, for his encouragement and support. We are also grateful to Messrs S. N’Gadjaga and P. A. Ki for computer programming and data processing, and to Messrs A. Daribi and R. Toe for editorial and cartographic assistance respectively. We also acknowledee the accurate work of the technicians of the EPI unit and the members of the National Teams of the seven OCP countries during data collection. References Ba, O., Karam, M., Remme, J. & Zerbo, G. (1987). Place des enfants dans l’evaluation du Programme de Lutte contre I’Onchocercose en Afrique de 1’Ouest. Tropical Medicine and Parasitology, 38, 137-142.

Erttmann, K. D., Meredith, S. E. O., Greene, B. M. & Unnasch, T. R. (1990). Isolation and characterization of form specific DNA ‘sequences of 0. volvulus. Acta Leidensia,

59, 253-260.

Dadzie, K. Y., Remme, J., Baker, R. H. A., Rolland, A. & Thylefors, B. (1990). Ocular onchocerciasis and intensity of infection in the community. III. West Africa rainforest foci of the vector Simulium sanctipauli. Tropical Medicine and Parasitology, 41, 371-450. De Sole, G. & Remme, J. (in press). Importance of migrants infected with Onchocerca volvulus in West African river valleys protected by 14 to 15 years of Simulium control. Tropical

Medicine

and Parasitology.

De Sole, G., Remmq, J. & Dadzie, K. Y. (1990). The epidemiological impact of vector control in the Onchocerciasis Control Programme in West Africa. I. Incidence and changes in prevalence and intensity of Onchocerca volvulus infection. Acta Leidensia, 59, 119126. Philippon, B., Remme, J., Walsh, J. F., Guillet, P. & Zerbo, G. (1990). Entomological results of vector control in the Onchocerciasis Control Programme in West Africa. Acta Leidensia, 59, 79-94. Plaisier, A. P., van Oortmarssen, G. J., Remme, J., Alley, E. S. & Habbema, J. D. F. (in press). The risk and dynamics of onchocerciasis recrudescence after cessation of vector control. Bulletin of the World Health Organization.

Prost, A. & Prod’hon, J. (1978). Le diagnostique parasitologique de l’onchocercose, revue critique des methodes en usage. Mkdecine Tropicale, 38, 519-532. Remme, J., Ba, O., Dadzie, K. Y. & Karam, M. (1986). A force-of-infection model for onchocerciasis and its applications in the epidemiological evaluation of the Onchocerciasis Control Programme in the Volta River Basin area. Bulletin of the World Health Organization, 64, 667-68 1. Remme, J., De Sole, G. & van Oortmarssen, G. J. (1990). The oredicted and observed decline in onchocerciasis infeccon during 14 years of successful control of Simulium spp. in West Africa. Bulletin of the World Health Organization, 68, 331-339. Walsh, J. F., Davies, J. B. & Le Berre, R. L. (1979). Entomological aspects of the first five years of the Onchocerciasis Control Programme in the Volta River Basin. Tropical Medicine and Parasitology, -30, 328-344. WHO (1987a). WHO Expert Committee on Onchocerciasis: third report. Geneva: World Health Organization, Technical Report series, no. 752. WHO (1987b). Twelve years of onchocerciasis control in western Africa. World Health Statistics Annual, 1%22. Received November

17 October 1990

1990;

accepted

for publication

28