AN EPIDEMIC OF URBAN PLAGUE IN MAJUNGA, W ESTERN COAST OF MADAGASCAR

In August 1991, an epidemic of urban plague broke out in Majunga, a po rt on the western coast of Madagascar. As the first resurgence of the disease in this area in 70 years, the episode was a dramatic reminder of the ubiquity of the disease and layed low the falsely reassuring id ea that plague was confined to well known centers of high altitude. Af ter the epidemic had claimed 12 victims, 21 patients suffering from bu bonic plague and 7 cases of pneumonic plague, it was brought under con trol by treating exposed subjects and introducing antivectorial measur es. Transmission to man continued in the shape of sporadic cases of bu bonic plague until April 1992 and a total number of 41 probable or con firmed bacteriological cases (not counting those deceased) were record ed out of 202 suspected cases. The history of the disease related almo st all these cases to the highly populated and unsanitary neighbourhoo d of Marolaka, situated on the edge of the Betsiboka estuary, and whic h has kept semi-rural characteristics. its murine population had pract ically disappeared, decimated by an epizootic in early July. Out of th e two captured sets of two Rattus rattus (41 fleas per animal) and two R. norvegicus each, three were carriers of the bacillus. Numerous shr ew-mice (Suncus murinus) occupied the zone and three tested positive t o the direct bacteriological test. For the insectivore, the mean Xenop sylla cheopis index was 4 (n=16) rising to 26 in positive animals. The discovery of a plague infested rat carrying X cheopis and the persist ence of P. irritans after insecticide treatments, led us to doubt the efficiency of the antivectorial fight that had bee undertaken. Risk of endemization was referred to. After an period of quiet of 3 years, is olated cases were detected in March and May 1995, followed by a new ep idemic between July 1995 and March 1996. 617 clinical suspects of bubo nic plague were notified. The confirmation by isolating the bacillus w as obtained for 60 subjects and 48 were considered to be probable case s. Twenty-four deaths were attributed to plague. Once again, the high density of cases came from Marolaka. Rats captured in lune 1995 reveal ed the presence of R. rattus and R. norvegicus in small numbers and of S. murinus carrying X. cheopis. In early November the Rattus had prac tically disappeared. There remained numerous S. murinus (n=84) with an X. cheopis index of 5, despite the pulverization of insecticides. The bacillus was isolated from R. Norvegicus as well as S. murinus. Plagu e is thus settling into the town of Majunga. The bacillus maintains it self in the estuary zone and in southern winters, for certain years, t ransmission among rodents is amplified, and then an epidemic breaks ou t. The epicenter of the disease is clearly Marolaka and its spread is limited to neighbouring areas. In such a biotype, the inter-epizootic cycle of the bacillus must be elucidated without delay. The actual rol e of shrew-mice as an efficient reservoir has not been established but quite obviously they allow the vectors to survive after the extinctio n of murine populations affected by plague. The operational efficiency of anti-vectorial measures should be evaluated in the field while tes ts on the sensitivity of fleas to insecticides have highlighted the ap pearance of new forms of resistance.