I wish to thank the British Society for Parasitologv for the privilege of being your Chairman this morning and for assembling such a promising series of papers for this Autumn Symposium.

Until recently, almost all studies of leishmaniasis epidemiology were qualitative and descriptive. But now that the natural history of many Leishmania parasites is quite well known, there is growing interest in quantitative analysis. In this paper I use mathematical models in conjunction with field data to try to answer a wider range of questions than has previously been possible with descriptive techniques, and to sharpen some of the outstanding questions for laboratory workers. This is done with reference to the persistence and resilience of canine leishmaniasis, the maintenance of virulence poly morphisms in Leishmania populations, and the possible existence of cycles of human kala-azar. I conclude by posing a set of problems under three headings: diagnosis of infection (as distinct from disease), natural immunity to Leishmania infection in the vertebrate host, and genetic variation in the parasite population. Some solutions from the laboratory can be found in the companion paper by Black (1992).

Application of quantitative methods to the study of leishmaniasis epidemiology has allowed Dye (1992) to pinpoint important biological parameters which, if they could be accurately measured in the field, would contribute most to our knowledge of the spread of disease and key targets for control. Three areas in which laboratory-based research could impact most on leishmaniasis epidemiology were highlighted by Dye (1992): (i) the development of accurate diagnostic tools which can distinguish between current and past infection; (ii) to determine the underlying molecular/genetic basis to virulence polymorphisms in the parasite and study these in the context of field epidemiological studies; and (iii) to provide the molecular tools to measure genetic variation in resistance to infection in humans and in reservoir hosts of disease. This paper describes current progress in attaining these goals, highlighting first the work on isolation and field application of genomic and kDNA probes for species-specific diagnosis, and the development of PCR-based assays which can be performed under field conditions. At a more preliminary stage, studies are described in which variability in the major molecular determinants of virulence (lipophosphoglycan, GP63, and members of the HSP7O family of stress proteins) identified through studies of laboratory models of infection, is being measured in primary field isolates of Leishmania peruviana. To complete the picture, current progress in identifying and cloning the genes which control host resistance to leishmanial infection is described, along with field studies of multicase families of human disease in which linkage analysis using marker genes from the chromosomal regions bearing these genes can be used to find evidence for their role in determining disease phenotypes in man. The leishmaniasis epidemiology will be all down to the DNA. projected view from these studies is that the future of leishmaniasis epidemiology will be all down to the DNA.

Molecular characterization of the Plasmodium falciparum genome has led to identification of polymorphic loci and the mechanisms generating genetic diversity in this parasite. This information has resulted in the development of molecular methods to type parasite diversity in the field. Consequently, we are now in a position to describe the population genetics and dynamics of P. falciparum. The limited number of field studies that have been conducted to date have revealed an extraordinary degree of genetic diversity in natural parasite populations. Heterozygous recombination which occurs during meiosis appears to be one mechanism for generating genetic diversity. The rate at which such recombination occurs in natural parasite populations defines the genetic structure of the parasite population and can influence the ability of the parasite to respond to selection pressure. The high frequency of single genotype infections and the female-biased gametocyte sex ratios found in hyperendemic malaria areas suggest that self-fertilization occurs frequently. Population- wide surveys of allele frequencies in endemic areas have, however, shown no evidence of linkage disequilibrium and are consistent with a panmictic population structure. We argue that these studies have only sampled symptomatic infections, within which rare or recombinant genotypes may be disproportionately represented. They also take no account of the spatial structure of P. falciparum populations. Systematic investigations of the amount of heterozygosity in small areas as part of population-wide surveys are required to define the genetic structure of P. falciparum populations. Population dynamic studies which consider genetic heterogeneity of P. falciparum have shown fluctuations of different serotypes in space and time. The host immune response appears to play an important role in generating these dynamics. Integrated field and laboratory studies, which consider the interaction between population genetics and dynamics, will be necessary to describe the population biology of P. falciparum.

In situations where malaria eradication is not an option in the foreseeable future the emphasis must be on the control of morbidity and mortality due to malaria. Under such circumstances drawing a distinction between malarial parasitization and malarial disease may be important for workers in both field and laboratory. This concept is explored from the points of view of the epidemiological picture of malaria in endemic populations, the factors which may influence progression to disease and the processes which mediate disease.

The basic tenet of the immunological perspective of fuiarial disease is that differential immune responsiveness among individuals exposed to infection results in the different clinical manifestations that develop. The mechanisms involved in this differential responsiveness appear to reflect different T-cell cytokine response patterns. Asymptomatic patients with the clinically silent presentation of ‘asymptomatic microfilaraemia’, who have been previously described as being ‘immunosuppressed’ with respect to their generating pro-inflammatory (Th1-type) immune responses to parasite antigen, are now recognized to be fully responsive to parasite antigen but to produce cytokines and mediators that have primarily anti-inflammatory (Th2-like) effects. Studies with immunodeficient mice have indicated the existence of two alternative pathways to the development of lymphatic pathology: one dependent on the induction of inflammatory reactions by the host immune response, the other entirely independent of the immune system and reflecting the direct actions of the parasite or its products on the lymphatics. As histopathology of affected human lymphatics is consistent with this hypothesis, it may be that the lymphatic pathology seen normally in the amicrofilaraemic, highly immunoresponsive infected patients derives from inflammation induced by immune responses to parasite antigen, whereas the lymphatic pathology sometimes seen coexisting with the ‘immunosuppressed’ state of asymptomatic microfilaraemia actually reflects lymphatic damage that is not immunologically mediated. Though little information exists about the ‘natural history’ of lymphatic filariasis, there is no evidence for an inevitable progression from one clinical form to another. Instead, there appears to be a definite plasticity in the response that depends on prior (? pre-natal) and current exposure to the parasite as well as on the immunomodulatory effects it induces. This plasticity does not appear to be complete, however, as there is no evidence that a chronically infected host who has developed strong pro-inflammatory immune responses can subsequently become sufficiently ‘down-regulated’ to support an asymptomatic microfilaraemia type of infection. Another possible constraint to the plasticity of the clinical and immunological responses may be the genetic determination of certain unusual syndromes, such as tropical pulmonary eosinophilia or TPE, though this hypothesis remains to be proven.

A major question in the study of any parasitic disease is the relationship between infection and clinical disease. The public health importance of lymphatic filariasis has generated a large body of research in this area, both in laboratory studies (Ottesen, 1984,1989; Maizels & Lawrence, 1991) and in the field (Hayashi, 1962; Hairston & Jachowski, 1968; Denham & McGreevy, 1977; Vanamail et al. 1989 b; Bundy, Grenfell & Rajagopalan, 1991; Srividya et al. 1991 b). Despite this, there is still no conclusive explanation for the apparently complex relationship between infection and clinical disease observed in human communities. At least part of the problem may lie in the current impossibility of measuring adult worm burden in vivo (Pichon et al. 1980; Denham & Fletcher, 1987; Das et al. 1990; Grenfell et al. 1990). Although there has recently been significant progress in the development of immuno logical markers for infection status in humans (Ottesen, 1989; Day et al. 1991 a), microfilaraemia is still the most reliable measure of current infection in the field. Studies in endemic areas indicate that, far from there being any simple direct relationship between microfilaraemia and disease status, it is possible to find some individuals with microfilariae in their blood but no disease, and indeed with all other combinations of infection and disease status (Hairston & de-Meillon, 1968; Hairston & Jachowski, 1968; Beaver, 1970; Bryan & Southgate, 1976; Denham & McGreevy, 1977; Pani et al. 1991). Furthermore, the proportions of people in different categories are often observed to vary between endemic areas (Denham & McGreevy, 1977; Day et al. 1991 a).

The mechanism by which small animals such as rodents resist or eliminate nematode parasites requires mucosal in flammation as the final effector of the immune response. The resulting freedom from chronic infection may be worth the price of short-term illness. Putative vaccines which attempt to enhance the natural effect will have to take into account the inflammatory cost to the host. Human helminthiases involve a more stable equilibrium between host and parasite. The medical literature on hookworm disease and clinical ascariasis describes, for the former, some chronic inflammatory effects correlated with worm burden, but for the latter a less quantified or predictable set of detrimental effects. We describe a current, systematic study of the inflammatory response to whipworm infection, in which anaemia, growth retardation and intestinal leakiness are viewed as predictable consequences related to infection intensity. There is evidence for the absence of cell-mediated immunopathology. However, a specific, IgE-mediated local anaphylaxis may, at least partly, mediate the deleterious effects. Increased numbers of mucosal macrophages may also contribute to the chronic, systemic effects through their output of cytokines. Similar attempts to show the mechanisms of pathogenesis and quantify the effects of hookworm disease should be undertaken.

The morbidity and transmission dynamics of geohelminthiases are determined by the patterns of infection intensity in the community. Understanding the determinants of these patterns requires a combination of field, laboratory and theoretical study. Studies of age-specific reinfection, and of the phenomenon of predisposition, indicate that the major determinant of convex age-intensity profiles and of heterogeneity in infection intensity is the rate of establishment of infection, rather than the rate of adult worm mortality. The rate of establishment is, in turn, determined by exposure to, and protection from, infection. The evidence indicates that exposure, at least to the orally-transmitted geohelminths, varies with age and is highly heterogeneous between hosts. The immune response in geohelminthiasis is vigorous, parasite-specific, hetero geneous between hosts, and both age and infection dose dependent, but has yet to be convincingly shown to be protective. Since the immune response is itself a function of exposure, unravelling the interaction between ecology and immunology as determinants of geohelminth worm burden will require simultaneous assessment of both processes via immuno epidemiological study.

Having been trained at the London School of Hygiene and Tropical Medicine, I am a tropical medicine man in the tradition of Sir Patrick Manson, that is, a parasitologist. In his address inaugurating the London School of Tropical Medicine in 1899, entitled The Need for Special Training in Tropical Medicine, Manson stated that ‘the peculiar dis tribution of a large class of tropical diseases depends, in the first place, on the fact that they are entozoal diseases, in the second place, that the entozoa concerned require intermediate and definitive hosts, and, in the third, that one or other of these hosts requires a high atmospheric temperature, in other words are native to warm climates.’ He went on, ‘today the protozoan and the helminth, as regards tropical pathology, are in the ascendant.’ This belief still holds sway, as the great British and American tropical medical journals remain largely devoted to parasitic infections.