Laboratory Experiments on the Control of Cyclops Transmitting Guinea Worm

Bull. Orjg. mond. Santc 1970, 42, 563-567 Bull. Wd Hith Org. Laboratory Experiments on the Control of Cyclops Transmitting Guinea Worm RALPH MULLER 1 Infection with the guinea worm (Dracunculus medinensis) affects millions of people living in rural areas in certain parts of the world. Usually, the standard of living is low and elaborate control measures would not be feasible. In this study, various insecticidal and molluscicidal compounds were tested in the laboratory for their toxicity to the intermediate hosts, namely, various species of cyclops, which often live in sources ofpotable water, such as step-wells, cisterns andponds. Cheapness, low toxicity to mammals and ease ofapplica- tion, in conjunction with effectiveness against cyclops, are primary requisites for any compound acceptedfor widespread use. Abate, an insecticide, has been shown to be parti- cularly promising in these respects. The timing and mode of application of compounds toxic to cyclops in various endemic regions are discussed. The disease dracunculiasis caused by the guinea worm (Dracunculus medinensis) affects millions of people in West and North Africa, the Middle East, Pakistan and India. It is transmitted by various species of carnivorous copepods (cyclops) living in small bodies of fresh water which often serve as sources of drinking-water, e.g., step-wells in India, cisterns (birkehs) in Iran and temporary ponds in most other endemic areas. The disease is confined to rural areas where the standard of living is low, which have no piped water supply, and in which expensive control measures would not be feasible. Control of the disease by treating the water to kill cyclops has often been advocated but rarely attempted, partly because of the cost but also because of a lack of the basic epidemiological knowledge necessary to make such treatment effective. Steam treatment of step-wells was sug- gested by Leiper (1911), the use of potassium permanganate by Turkhud (1919) and of quicklime and slaked lime by Davis (1931). More recently, Ramakrishnan & Rathnaswamy (1953) found in the laboratory that 10 ppm of DDT caused 100% mortality of cyclops in 48 hours, and using a single application of DDT at this dilution, Nugent, Scott & Waddy (1955) reduced the incidence of dracuncu- 1 Lecturer, Department of Medical Helminthology, London School of Hygiene and Tropical Medicine, London, England. liasis markedly (from 26.50% to 6%) in 5 out of 7 villages in an area of Ghana. These authors believed that the lack of success in the other two villages was due to faulty technique. The most recent attempt to control the disease by chemical means is in the south of Iran, where chlorination of the birkehs (combined with measures designed to prevent guinea worm larvae from reaching the water) has led to its almost complete eradication (Sabokbar, 1968). The success of the campaign can be judged from the fact that, according to the medical officer in charge of the Government Health Clinic at Bastak (a town in the centre of the endemic area), in the two years previous to 1967 there had been only one case in the town. Lindberg (1936) estimated that 15 %-20% of the population was infected annually. This improvement occurred des- pite the fact that domestic water is still obtained from birkehs. In recent years, many compounds have been discovered that are effective as molluscicides or insect larvicides and it was thought that some might also be effective against cyclops. Most of these substances are likely to be inexpensive in the quantities required and to be available in suitable formulations. Their toxicity to mammals has been studied. Gretillat (1965) has already shown in the laboratory that the molluscicide zinc dimethyldithio- carbamate (ziram) is effective at a concentration of 0.25 ppm for 24 hours. 2497 -563- 564 R. MULLER MATERIALS AND METHODS In order to obtain consistent experimental results, it is desirable to maintain colonies of cyclops in the laboratory. The first species used was Cyclops leuckarti (Claus), a colony of which was set up from specimens obtained from an infected pond in Western Nigeria. This is one of the most important species in the transmission of guinea worm, acting as an intermediate host in all endemic regions investigated, but this species is not easy to keep in the laboratory in large numbers. For this reason, all the tests reported here were carried out with specimens of C. vernalis (Fischer). The original stock of this species was obtained from a pond in Regent's Park in London. C. vernalis occurs in small ponds in many endemic areas and, although it has not been reported as a host in nature, our studies have shown that it is very susceptible to experimental infection with guinea worm in the laboratory. Larvae develop normally in experi- mental infections and reach the infective stage in 12-14 days at 24°C, as in C. leuckarti. The few pre- liminary tests carried out with C. leuckarti showed that both species of cyclops had a similar range of susceptibility, at least to the chlorinated hydro- carbons. Cyclops could be reared in large numbers in 50-litre polyethylene tanks containing water main- tained at 24°C. The water used for filling the tanks was preconditioned by keeping a stock of guppies (Lebistes reticulatus) in it; this treatment ensured that dechlorination was complete. The cyclops were fed twice a week on hay infusion, as described by Mueller (1959). When the contents of a tank were strained through a sieve of suitable (210 ,u) mesh, the nauplii larvae of the cyclops could pass through and thus maintain the colony. If the nauplii larvae were required for testing, they could be separated by a sieve of 150-,i mesh. The copepodid stages could be separated from the adults by straining the water through a sieve of 250-,t mesh. In addition to the tests on the laboratory-reared cyclops, all the more promising compounds were tested against cyclops obtained from the field. This was done to eliminate results that might be due to the catastrophic fall in numbers that periodically affects laboratory cultures of invertebrates. Tests were carried out in Petri dishes containing batches of 5 adult cyclops in 10 ml of solution, and kept at a temperature of 24°C. Mortality was determined after 24 hours under a stereoscopic microscope. Mortality in the controls was some- times as high as 10% because of predation by the cyclops, but this could be allowed for by means of Abbott's formula. Care was necessary in trans- ferring cyclops by pipette, and any found dead or injured at the start of an experiment were replaced. Initial tests were carried out at concentrations of 2.5 ppm, 0.5 ppm and 0.1 ppm in pond water, with 5 Petri dishes for each concentration and 10 con- trols. Compounds which caused 100% mortality at 0.1 ppm were tested over a narrower range of concentrations to give a mortality-dose curve, from which the the LC50 and LCgo were obtained. All compounds were tested on two separate occasions, and in the secondary series some have been tested up to 10 times. The compounds which have been tested are shown in Table 1. RESULTS The results of the primary screening programme are given in Table 1, and of the secondary screening programme in Table 2. The results of the tests using adult cyclops only are recorded since the copepodid and nauplius stages were more sensitive to all the compounds tested. No tests were con- ducted on eggs of cyclops. In general, insecticides were more effective than molluscicides, and the chlorinated hydrocarbons and some of the organophosphorus compounds were outstanding. Of the organophosphorus compounds that caused 100% mortality at a concentration of 0.1 ppm., Abate (0, 0, O', O'-tetramethyl 0, 0'- thiodi- p-phenylene phosphorothioate) 1 appeared to show the most promise. This compound has been used extensively in potable water supplies as a mosquito larvicide (Brooks, Schoof & Smith, 1965; Schoof, 1967; Laws et al., 1968) and has a very low mam- malian toxicity. The other compounds listed in Table 2 may be equally, if not more, effective in the field, but there is less information available on their suitability for use in treating potable water. Further tests carried out with Abate showed that the 50% emulsifiable concentrate formulation gave very variable results but the 20% and 50% wettable powders and the 1% sand-granule formulations were found to be as effective as the technical product. The 1% sand-granule formulation is known to have a more prolonged residual effect than the other formulations and determinations were made of the 1 Produced by the Union Carbide Company. EXPERIMENTS ON THE CONTROL OF CYCLOPS TRANSMITTING GUINEA WORM 565 TABLE 1 PRIMARY SCREENING OF COMPOUNDS FOR ACTIVITY AGAINST CYCLOPS UNDER LABORATORY CONDITIONS Minimum Maximum concen- concen- Compound (and supplier) OMS tration tration No. causing causing100% no mortality mortality(ppm) (ppm) DDT 16 2.5 - dieldrin 18 0.1 gamma-HCH 17 0.1 ziram - 2.5 0.1 dichlorvos (Shell Chemical Co.) 14 0.1 WL 12730 (Shell Chemical Co.) a - 0.5 -0.1 SD 8447 - 0.5 <0.1 (Shell Chemical Co.) a chlorfenvinphos(Shell Chemical Co.) - 0.5 0.1 niclosamide ethanolamine salt (Bayer Chemical Co.) a _ 2.5 <0.5 fentin acetate (Pure Chemical Ltd.) a - 2.5 0.5 sodium pentachlorophenate(Dow Chemical Co.) a - >2.5 <0.1 fenthion (Bayer Chemical Co.) 2 2.5 0.1 diazinon (J. R. Geigy, A. G.) 496 >2.5 0.5 malathion (American Cyanamid Co.) 1 0.5 0.1 Molucid (ICI 24223)(Imperial Chemical Industries Ltd.) a - 2.5 0.1 azinphosmethyl (Bayer Chemical Co.) - 0.1 - dimethan (J. R. Geigy, A. G.) - 0.1 propoxur (o-isopropoxy- phenyl methylcarbamate)(Bayer Chemical Co.) 33 0.1 Pyrolan (J. R. Geigy, A. G.) 20 0.1 fenitrothion 43 0.1 (Bayer Chemical Co.) Abate (Union Carbide Corp.) 786 0.1 Dursban(Dow Chemical Co.) 971 0.1 - calcium hypochlorite - 2.5 0.5 methidathion (J. R. Geigy, A. G.) - 0.1 - a Molluscicide. TABLE 2 SECONDARY SCREENING OF SELECTED COMPOUNDS FOR ACTIVITY AGAINST CYCLOPS UNDER LABORATORY CONDITIONS Compound LCso LCso dichlorvos 0.012 0.018 Pyrolan 0.001 0.008 fenitrothion 0.0008 0.08 Abate 0.002 0.006 Dursban 0.00001 0.0008 duration of the residual effect. Each of 12 small aquaria was filled with 2 litres of water and kept in sunlight at 24°C. The sand-granule formulation of Abate was added to 10 of the aquaria to give a concentration of 1 ppm of active compound; the other 2 aquaria were kept as controls. About 100 cyclops were added to each tank at 3-weekly intervals and the mortality was assessed after 24 hours. Under these conditions, there was 100% mortality of the cyclops in the test tanks over a period of 12 weeks, in spite of the fact that the sides of the tanks rapidly became covered with green algae. This can be compared with the effectiveness of Abate in various formulations and concentrations against the larvae of Aedes aegyptii: more than 5 weeks effectiveness in water-storage tanks in the American Virgin Islands using a sand-granule formulation at a concentration of 0.25 ppm (Brooks, Schoof& Smith, 1965), satisfactory kills for 15 weeks in 210-litre oil-drums with a concentration of 1.0 ppm in an emulsion formulation (Schoof, 1967) and more than 2 months effectiveness in trials in Thailand with a concentration of 1.0 ppm in a sand-granule formulation.' A field trial was also undertaken in a pond in an endemic region of Western Nigeria. The pond was in a shaded situation and the population of cyclops was estimated by taking samples with a plankton net at the edge and at various depths in the middle; sampling was always carried out at the same time of the day. Abate (technical product) at a con- ' Bang, Y. H. & Tonn, R. J. (1969) Evaluation of 1 % Abate (OMS-786) sand granules for the control of Aedes aegyptii larvae in potable water; unpublished WHO working document VBC/69.121. A limited number of copies of this document is available to persons officially or professionally interested on request to Distribution and Sales, World Health Organization, 1211 Geneva, Switzerland. R. MULLER centration of 0.6 ppm was added to the pond and subsequently no cyclops were found for 6 weeks. Ostracods and tadpoles of Xenopus muelleri were unaffected. DISCUSSION The results obtained with Abate are at variance with those obtained by Ruber & Baskar (1968), who found in the laboratory that a salt-marsh species, Cyclops spartinus, was unaffected by Abate up to a concentration of 1 ppm. They found Dursban (see Table 1) to be rather more effective than Abate (LC,, of 0.1 ppm), although not nearly so potent as in the present tests. A diaptomid crustacean was found to be rather more susceptible to Abate (LC,o of 0.002 ppm) than the cyclops. The variations may reflect specific differences in susceptibility to Abate or may possibly be due to the different ecological habitats. The results obtained with chlorinated hydro- carbons were similar to those of other investigators (Ramakrishnan & Rathnaswamy, 1953; Ruber, 1967). It is probable that many compounds in addition to those tested, particularly some new insect larvi- cides, would kill cyclops at low concentrations. However, it is important that any compound added to potable water sources to control cyclops should not affect the quality of the water, otherwise its application will be vigorously opposed by the consumers. A recent attempt to chlorinate step- wells in the Udaipur region of India had to be abandoned when local farmers protested that their cattle would not drink the water for many days afterwards, although all taste appeared to have vanished (Banks, personal communication). Were it not for its cumulative toxicity to other forms of life, DDT would probably be the most satisfactory compound for controlling cyclops on a cost-effectiveness basis; DDT has the added advan- tage that its effectiveness in the field has already been demonstrated. It is not recommended, however, that this class of compound should be added to drinking-water. The timing of applications must be carefully considered, taking into consideration local patterns of transmission, since the epidemiological pattern varies greatly in different endemic areas. In Africa, infection occurs almost entirely from ponds; mainly during the rainy seasons in desert and savanna regions, while in the forest and derived savanna areas of West Africa transmission takes place primarily in the dry season. In the forest and derived savanna of south- western Nigeria the rains from June to September are so heavy that many ponds turn into streams and the cyclops are washed away; even in those ponds which remain isolated, the density of cyclops is greatly reduced and the numbers are adversely affected by the turbidity of the water. There appears to be little transmission during this time but, as the ponds are never completely empty during the dry season, some transmission occurs in every month from October to May. The peak incidence of infection in the human population is usually during March and April when the ponds are at their lowest level. An example of the infection pattern found in a desert area is provided by the focus in the Sind desert in the Hyderabad region of Pakistan. For most of the year water is taken from deep draw-wells, of which there may be up to 20 in a village. These wells generally have a parapet about a metre high surrounding them and they seem unlikely to be of importance in the transmission of the disease. However, during the rains (July-September) one or two very large ponds, known as tarais, are formed in each village and many villagers take water from them, partly because the water in the wells is slightly brackish and partly because a charge is made for the collection and distribution of well water; there are also religious reasons. The period of the year when worms emerge corresponds closely to the months when there was water in the tarais the previous year (Ansari & Nasir, 1963). It is evident that the majority of infections, if not all, are derived from these ponds. The number of applications of insecticide neces- sary to control or eliminate dracunculiasis in an area will depend on the residual activity of the com- pound. However, in areas with a transmission season as short as that in the Sind desert 2 applica- tions per year should be sufficient, and control can be effective even with the use of a compound, such as bleaching powder, having no residual activity. A first application could be made 4 weeks after the ponds refill, and a second application 4 weeks later if necessary (the second application need only be made if sampling shows a reappear- ance of cyclops). This timing is based on two assumptions: first, that it will be at least 2 weeks after the pond has refilled (or after treatment) before there are enough cyclops present for transmission 566 EXPERIMENTS ON THE CON'IROL OF CYCLOPS TRANSMITTING GUINEA WORM 567 to occur; second, that the larvae take a minimum of 12-14 days to reach the infective stage (Fairley & Liston, 1924; Onabamiro, 1958; Muller, 1968). In areas with a long transmission season, or where there is continuous transmission (as in the open step- wells in India and the birkehs in Iran) a compound with a long residual activity would be of great ad- vantage, particularly where there are many scat- tered water sources. The present study has indicated that any of the following would be effective: Abate (particularly the sand-granule formulation), dich- lorvos, Dursban, fenitrothion and Pyrolan (see Tables 1 and 2). ACKNOWLEDGEMENTS This work was supported by a grant from the Ministry of Overseas Development and the Medical Research Council of Great Britain. Invaluable technical assistance was provided by Miss P. Nduru and Mrs V. Duckett. I am grateful to Dr A. Smith and staff of the British Museum (Natural History) for the identification of Cyclops species, and to the manufacturers of the various compounds tested for donating samples of their products. RtSUME EXPtRIENCES DE LABORATOIRE RELATIVES A LA LUTTE CONTRE LES CYCLOPES, HOTES INTERMEDIAIRES DU VER DE GUINtE Une vingtaine de composes chimiques ont fait l'objet, au laboratoire, d'essais d'activitd contre les cyclopes (Cyclops vernalis) hotes intermediaires du ver de Guinee (Dracunculus medinensis), agent de la dracunculose. Apres des tests pr6liminaires utilisant des concentrations de 2,5, 0,5 et 0,1 parties par million, on a soumis les com- poses les plus actifs (c'est-a-dire tuant 100% des cope- podes a la dose de 0,1 partie par million) a des ipreuves plus poussees avec une gamme moins etendue de concen- trations, etabli des courbes dose-mortalite et determine les CL5* et les CL1. Dans l'ensemble, les insecticides - notamment les hydrocarbures chlores et certains organophosphores, qui font preuve d'une efficacite remarquable- se sont r6v616s plus actifs que les molluscicides. Si l'on tient compte de facteurs comme le faible prix de revient, la facilit6 d'appli- cation et le peu de toxicite pour les mammif6res, lAbate, de'js largement employe dans la lutte contre lee larves d'insectes, offre des avantages appreciables en tant que moyen de destruction des cyclopes. D'autres compoe's, comme le dichlorvos, le Dursban, le fenitrothion ct le Pyrolan pourraient aussi etre utilises avec succes. L'auteur examine dans quelle mesure les donn6ees pi- demiologiques sont susceptibles d'influencer lee moda- lites de la lutte chimique contre les cyclopes dans diverses regions d'enddmicite. REFERENCES Ansari, A. R. & Nasir, A. S. (1963) Pak. J. Hlth, 13, 152-167 Brooks, G. D., Schoof, H. F. & Smith, E. A. (1965) Mosquito News, 25, 423-427 Davis, J. L. (1931) Trans. roy. Soc. trop. Med. Hyg., 24, 631-633 Fairley, N. H. & Liston, W. G. (1924) Indian J. med. Res., 12, 93-103 Gretillat, S. (1965) Biol. med. (Paris), 54, 529-539 Laws, E. R. et al. (1968) Bull. Wld Hlth Org., 38, 439-445 Leiper. R. T. L. (1911) J. Lond. School trop. Med., 1, 28-30 Lindberg, K. (1936) Arch. Schiffs- u. Tropenhyg., 40, 330-341 Mueller, J. F. (1959) Trans. Amer. microsc. Soc., 78, 245-255 Muller, R. (1968) J. Helminth., 32, 331-338 Nugent, D. A. W., Scott, D. & Waddy, B. B. (1955) Trans. roy. Soc. trop. Med. Hyg., 49, 476-477 Onabamiro, S. D. (1958) Ann. trop. Med. Parasit., 50, 157-166 Ramakrishnan, N. R. & Rathnaswamy, G. K. (1953) Indian med. Gaz., 88, 386-390 Ruber, E. (1967) Proc. New Jersey Mosq. Exterm. Assoc., 54th Ann. Mtg, pp. 139-144 Ruber, E. & Baskar, J. (1968) New Jersey Mosq. Exterm. Assoc., 54th Ann. Mtg, pp. 99-103 Sabokbar, R. (1968) Dracunculose en Iran. In: Abstracts and Reviews of the 8th International Congresses on Tropical Medicine and Malaria, Teheran, p. 938 Schoof, H. F. (1967) Bull. WId Hlth Org., 36, 618- 622 Turkhud, D. A. (1919) Indian J. med. Res. (Special Indian Science Congress Number), p. 217