Bull. Org. mond. Sante 1967, 36, 533-535 Bull. Wld Hlth Org.

Aedes aegypti and Other in Relation to the Dengue Syndrome

P. F. MATTINGLY 1

During the past 13 years there have been four introduction of a wild virus into the human popula- major outbreaks of new or previously unrecognized tion, to the appearance of a radically new virus. human -borne virus diseases in Asia and No one explanation is likely to fit all four cases, Africa. The diseases in question are but a common factor may perhaps be found in fever, o'nyong nyong fever, haemorrhagic dengue human interference with the biological environment. and Kyasanur Forest disease. It is possible that there may have been previous, unrecognized epi- THE SEMLIKI FOREST SUBGROUP OF GROUP-A demics of chikungunya fever, which is clinically ARBOVIRUSES indistinguishable from true dengue, but this has not been proved. O'nyong nyong is unique among all This subgroup includes three African viruses: the known human arboviruses in having as its Semliki Forest, chikungunya and o'nyong nyong. vectors anopheline mosquitos. Bearing this in mind, All three have antigenic components in common, as well as the fact that it spread over a vast area of Semliki Forest being the most inclusive in this Africa and involved more than a million human respect and o'nyong nyong the least (Porterfield, cases, it is difficult to believe that it is not a radically 1961). On the assumption that they are phylogene- new disease. There is as yet no general agreement tically related, it would therefore be reasonable to as to whether haemorrhagic dengue should be derive chikungunya from Semliki Forest and regarded as a distinct clinicalentity. Majority opinion o'nyong nyong from chikungunya. This is almost (gleaned at first hand during a recent visit to various certainly on over-simplified picture, but it may serve parts of Asia) seems, however, to incline to the view as a rough guide to relationships. that it represents one end of a continuous spectrum, The has been isolated from with the milder forms of true dengue at the other Aedes (Aedimorphus) and Eretmapodites spp. (Had- end. Kyasanur Forest disease is essentially a disease dow, 1960; McIntosh, Worth & Kokernot, 1961) in of forest monkeys communicated to man as a zoo- Uganda and Mozambique, and from a mixed pool of nosis. Available records give no indication of its Culex (Culex) and Aedes (Neomelaniconion) spp. in previous occurrence in India as an important human Uganda. The occurrence of immunity in forest disease. Its vectors are ticks and, possibly, to a small monkeys suggests that arboreal mosquitos may also extent, lice. These are relatively sessile vectors be involved (Haddow, 1960). This virus has been (Mattingly, 1965) and its ecology is likely, therefore, transmitted successfully in the laboratory, using to differ in important respects from that of the other A. aegypti, A. (Finlaya) togoi and a variety of diseases mentioned, which are -borne. Anopheles spp. (Collins, Harrison & Skinner, 1964). It will not be further discussed here, but it is sug- Chikungunya virus was isolated from pools of gested that some of the general principles which are A. aegypti, Culex spp. and Anopheles spp. during here invoked might be borne in mind when con- the original outbreak (Lumsden, 1955). It was sidering its possible origins. subsequently isolated from Aedes (Stegomyia) africa- In searching for the origins of these outbreaks, nus (Weinbren, Haddow & Williams, 1958) and, in a variety of possible causes must be taken into Asia, from Mansonia spp., A. aegypti and C. fatigans account, ranging from a simple increase in patho- (Rao, 1964). During an epidemic in Southern genicity of a pre-existing human virus, through the Rhodesia, A. (Diceromyia) spp. were shown to be efficient laboratory vectors; indeed, one of these was responsible for an I Principal Scientific Officer, British Museum (Natural accidental laboratory trans- History), London, England. mission to man (Paterson & McIntosh, 1964). 1974 -533- 534 P. F. MATTINGLY

The only known vectors of the o'nyong nyong perennial problem in arbovirus studies and has virus are Anopheles spp., mainly Anopheles funestus nearly always proved baffling. There are, however, and, to some extent, Anopheles gambiae; there can again some interesting coincidences. The distance be no doubt that these were the vectors responsible from the Mazoe Valley to Newala is about 700 miles for a recent epidemic in man (Williams et al., 1965). (1120 km), from Newala to the Pare-Taveta area The occurrence, for the first time in recorded about 450 miles (730 km) and from the Pare-Taveta history, of a viral disease of man, with anopheline area to the site of the o'nyong nyong outbreak about mosquitos as the major vectors, is a very remarkable 650 miles (1050 km). The intervals between the event and one that can hardly fail to cause disquiet, initiation of spraying and the first detection of virus especially as it coincided with a period of extensive were about 3, 21/2, and 31/2 years, respectively. activity against anophelines, associated with various It does not appear, therefore, that there would be malaria-eradication campaigns. While mentioning any need to account for any great discrepancies in the possibility of a causal connexion, Reeves (1965) the rate of travel during this period. More impressive considers that there is no evidence for it. He may is the fact that, after this route had been predicted by be correct, but there are some very remarkable the author in 1960, it became clear that o'nyong coincidences, and it would seem that these should be nyong virus was spreading southwards, along a very taken into account when considering other and similar route, in the human population. It subse- comparable outbreaks, e.g., haemorrhagic dengue. quently moved southwards along the east side of The two antimalaria projects most closely related, Lake Victoria (with the establishment of a secondary geographically and chronologically, to the initial centre to the west of the lake, which is probably not outbreaks of chikungunya and o'nyong nyong are significant) and entered Tanzania (Williams, Woodall the Mazoe Valley project in Southern Rhodesia & Gillett, 1965). It then progressed as far as southern (Alves & Blair, 1953) and the Pare-Taveta scheme Malawi, with offshoots reaching the Kenyan and in Tanzania (Wilson, 1960), respectively. The most Tanzanian coasts (Woodall, Williams & Lule, probable basis for any causal connexion would 1964). seem to be a deviation of Anopheles to an enzootic The evidence, such as it is, seems less convincing reservoir (possibly Semliki Forest) in consequence in the case of chikungunya than in the case of of spraying. It is therefore possibly significant that o'nyong nyong, more particularly as attempts to the best available evidence for such variation from transmit the former with anopheline mosquitos in Africa comes precisely from the Mazoe Valley (with the laboratory have so far failed. So far as is known, Swaziland) and the Pare-Taveta area. It is now however, only Anopheles gambiae and Anopheles known that, in the Mazoe Valley, anthropophilic stephensi have been tried, and it would seem desir- Anopheles gambiae A and B were wholly or largely able to test members of the Anopheles funestus and eradicated, leaving only zoophilic Anopheles gam- possibly other members of the Anopheles gambiae biae C (Paterson, 1963; Davidson, 1964). There complex (Paterson & McIntosh, 1964; Rao, 1964). may well also have been a replacement of Anopheles funestus by other, zoophilic members of the Ano- pheles funestus complex. Anopheles funestus is be- HAEMORRHAGIC DENGUE lieved to have been greatly reduced, but there are The term " haemorrhagic dengue " is preferred several other members of the complex in the area. to " haemorrhagic fever ", first, because the causal In the Pare-Taveta area, there is unequivocal agent belongs to the true dengue complex and, evidence of a massive replacement of Anopheles secondly, because the term " haemorrhagic fever " funestus by zoophilic Anopheles rivulorum (Gillies & has been used widely with reference to other, un- Smith, 1960). related fevers. Chikungunya virus has been several The next requirement would be for a transfer of times associated with outbreaks of haemorrhagic modified virus from the Mazoe Valley to the Newala dengue but has not, as far as is known, been shown area, where chikungunya disease was first detected, to be associated with the severe haemorrhagic syn- and subsequently into the Pare-Taveta area and the drome. The group-B arboviruses causing true dengue area in north-west Uganda where the first outbreak form a complex with some half-dozen serologically of o'nyong nyong disease occurred. As to the distinguishable members (dengue 1, 2, 3, 4 etc.). means by which such transfer may have been For present purposes, however, they may be treated accomplished, there is no suggestion here. This is a as a single entity, since too little is known regarding A. AEGYPTI AND OTHER MOSQUITOS IN RELATION TO DENGUE SYNDROME 535

differences in their ecology, vector relationships or was also able to demonstrate, experimentally, that clinical effects for any useful purpose to be served A. aegypti has a perceptible advantage when the by separating them. two species are reared together, either in artificial The only vectors so far conclusively implicated containers or in tree-holes. This might be expected are Aedes (Stegomyja) spp. Members of the A. (St.) to lead to a gradual replacement of A. albopictus, scutellaris complex may be responsible for transmis- although I would suspect that, in rural areas in sion in the Pacific area, but the only species known which natural habitats were relatively abundant, the to be important in the areas affected by haemorrhagic process would be a slow one. " Urbanization " of dengue are A. aegypti and A. albopictus (Rao, 1964). any kind (e.g., increased availability of domestic A. aegypti occurs as a wild mosquito only in Africa, water or increased use of artificial containers) would, and its occurrence elsewhere in the world is in all on the other hand, tend to accelerate it. Smith cases essentially attributable to human introduction (1956) considers that dengue has long been present (Mattingly, 1957). A. albopictus is essentially an in an endemic form in rural areas of tropical Asia. oriental mosquito with a limited extension into the He suggests that invasion of such areas of endemicity Australasian region and as far west as the Seychelles by A. aegypti could well lead to the establishment of and Comoro Island groups. Macdonald (1956a) more virulent strains of the virus. This is so much has produced evidence to show that A. aegypti was in line with the conclusions of the author regarding probably confined to coastal areas in Malaya up the origins of both o'nyong nyong fever and hae- to about 1900, and that it has been slowly encroach- morrhagic dengue that it seems that his words may ing into inland areas since that time, the process have been prophetic. If the complex etiology of being still far from complete. Smith (1956) has dengue in Asia is to be understood, it is essential brought forward further evidence suggesting pro- that equal attention be paid to both major vectors. gressive replacement of A. albopictus by A. aegypti Even the meagre knowledge of the role of A. aegypti elsewhere in South-East Asia. Macdonald (1956b) far outweighs the knowledge of the role of A. albo- found that A. aegypti was relatively more abundant pictus in rural areas. Common prudence suggests in built-up areas and A. albopictus in kampongs. He that this is a situation that urgently needs remedying.

REFERENCES

Alves, W. & Blair, D. M. (1953) Trans. roy. Soc. trop. Paterson, H. E. (1963) S. Afr. J. med. Sci., 28, 33 Med. Hyg., 47, 299 Paterson, H. E. & McIntosh, B. M. (1964) Ann. trop. Collins, W. E., Harrison, A. J. & Skinner, J. C. (1964) Med. Parasit., 58, 52 Mosquito News, 25, 54 Porterfield, J. S. (1961) Bull. Wid Hlth Org., 24, 735 Davidson, G. (1964) Riv. Malar., 43, 167 Rao, T. R. (1964) Indian J. med. Res., 52, 719 Gillies, M. T. & Smith, A. (1960) Bull. ent. Res., 51, 243 Reeves, W. C. (1965) Ann. Rev. Ent., 10, 25 Haddow, A. J. (1960) Bull. ent. Res., 50, 759 Smith, C. E. G. (1956) J. trop. Med. Hyg., 59, 243 Weinbren, M. Lumsden, W. H. R. (1955) Trans. roy. Soc. trop. Med. P., Haddow, A. J. & Williams, M. C. Hyg., 49, 33 (1958) Trans. roy. Soc. trop. Med. Hyg., 52, 253 Williams, M. C., Woodall, J. P., Corbet, P. S. & Gillett, Macdonald, W. W. (1956a) Ann. trop. Med. Parasit., 50, J. D. (1965) Trans. roy. Soc. trop. Med. Hyg., 59, 300 385 Williams, M. C., Woodall, J. P. & Gillett, J. D. (1965) Macdonald, W. W. (1956b) Ann. trop. Med. Parasit., 50, Trans. roy. Soc. trop. Med. Hyg., 59, 186 399 Wilson, D. B., ed. (1960) Report on the Pare-Taveta McIntosh, B. M., Worth, C. B. & Kokernot, R. H. (1961) Malaria Scheme 1954-1959, Dar-es-Salaam Trans. roy. Soc. trop. Med. IIvg., 55, 192 Woodall, J. P., Williams, M. C. & Lule, M. (1964) Mattingly, P. F. (1957) Ann. trop. Med. Parasit., 51, 392 Annual Report of the East African Virus Research Mattingly, P. F. (1965) New Scientist, 25, 22 Institute, 1962-1963, Entebbe, p. 31