42 African trypanosomosis

The African trypanosomoses are diseases of both man and his livestock. There are two forms of human trypanosomosis or sleeping sickness: Gambian or Rhodesian sleeping sickness, roughly corresponding to a West, Central or East African distribution respectively. Gambian sleeping sickness runs a more protracted and chronic course than the Rhodesian form; nevertheless, human trypanosomosis is invariably fatal if not treated. Animal reservoir hosts, both wild and domestic, assume greater importance for Rhodesian sleeping sickness than Gambian sleeping sickness, and the former is often an occupational hazard of those visiting or working in wildlife areas, e.g. tourists, hunters. Animal trypanosomosis transmitted by tsetse is generally referred to as Nagana, while the disease transmitted by other bloodsucking flies outside the African tsetse belt is known chiefly as Surra, but also by a variety of local names.

Trypanosomosis is caused by obligate parasitic protozoan flagellates of genus Trypanosoma. Trypanosoma brucei gambiense and T. b. rhodesiense are the causative organisms of Gambian and Rhodesian human trypanosomosis respectively. The third subspecies T. b. brucei is not infective to man. All 3 subspecies are morphologically indistinguishable. Several other trypanosome species cause Nagana besides T. b. brucei, but none is normally infective to man. Surra is caused by T. evansi, which is a mechanically-transmitted, mutant derivative of T. b. brucei and likewise not normally infective to man.

About 250 endemic foci of sleeping sickness are spread through 36 countries in sub-Saharan Africa and 60 million are estimated to be at risk. The annual incidence is currently in the tens of thousands and thought to be falling, but accurate figures are difficult to obtain especially from remote or inaccessible foci. Epidemics flare up in the endemic foci usually as a result of failure of control measures; however, new outbreaks can also occur in areas of resettlement and have also been linked to livestock movement. Sleeping sickness is largely restricted to rural areas by the distribution of its vector, the tsetse fly; both the fly and animal trypanosomosis have a far wider distribution than the human disease and Surra has spread to regions such as the Middle East, South America and Asia, way beyond the confines of the tsetse belt.

Sleeping sickness is transmitted by an insect vector, the tsetse fly (genus Glossina). Several species of the palpalis and morsitans groups are involved, which live in a variety of habitats and show different host preferences. Adults of both sexes are blood feeders and thereby transmit trypanosomes. The trypanosomes undergo a multiplicative and developmental cycle in the fly, taking a minimum of two weeks for appearance of infective forms. Individual flies can remain infective all their lifespan—several months under favourable conditions. Other routes of transmisson are relatively unimportant in the epidemiology of the tsetse-transmitted species, e.g. via infected blood or raw meat, mechanically vector borne, but are the main transmission routes for T. evansi.

Sleeping sickness control measures are aimed either at the trypanosome or the fly. Human cases are detected by active or passive surveillance and cured by treatment with trypanocidal drugs. Control of the tsetse vector is by application of residual insecticides or bush clearing and, more recently, by traps or insecticide-impregnated targets, or by wholesale release of sterile males. Tsetse control is more widely employed for the control of animal trypanosomosis than sleeping sickness.

Sleeping sickness aroused the curiosity and wonder of visitors to tropical Africa long before its cause was discovered, because of the extraordinary comatose state characteristic of the late stage of the disease. The first mention of sleeping sickness is by a fourteenth century Arab scholar, al-Qalqashandi (Hoare 1972). In the eighteenth and nineteenth centuries, European slave traders working on the West African coast were sufficiently aware of the disease to avoid buying Africans showing enlarged cervical lymph nodes, an early sign of the disease. Despite such precautions many people in the early stages of sleeping sickness were transported to the West Indies during the years of the slave trade, but the disease never became established there, presumably because a suitable vector was lacking.

Various causes for sleeping sickness were put forward, mostly relating to the African way of life, since Europeans appeared not to be susceptible. However, the scientific search for a causative agent became imperative at the turn of the twentieth century as a series of major epidemics began to threaten European trade and colonization in Africa. At the request of the British Foreign Office, the Royal Society sent two Sleeping Sickness Commissions to investigate a major epidemic in Uganda. The second commission under the leadership of David Bruce, swiftly revealed the cause of the disease, a discovery considered to be one of the greatest triumphs of tropical medicine, which was also sadly not without controversy (Davies 1962).

Trypanosoma gambiense Dutton, 1902 was initially described from the blood of a patient with ‘trypanosoma fever’, a mild disease of the West African Coast (Hoare 1972). Soon after this, Castellani, a member of the Royal Society Commission on sleeping sickness in Uganda, found trypanosomes in the CSF of sleeping sickness patients, and suggested trypanosoma fever to be the early stage of sleeping sickness, before the trypanosomes had infiltrated the central nervous system (CNS) from the bloodstream. Bruce and his wife, fresh from their discovery in South Africa that tsetse flies transmitted the pathogenic trypanosomes which caused Nagana in livestock, quickly grasped the significance of Castellani’s findings and demonstrated that the vector of sleeping sickness was a tsetse fly. These results were published in 1903.

Subsequently, a second human trypanosome was isolated in Zambia and named as a distinct species, T. rhodesiense Stephens and Fantham, 1910, on the basis of its morphology and virulence to experimental rodents. Cases of Rhodesian sleeping sickness were sporadic at first, but once the disease assumed epidemic form, its acute and rapid nature compared to Gambian sleeping sickness was appreciated.

The pathogenesis of sleeping sickness was greatly clarified once it was discovered to be caused by a tsetse-transmitted protozoan parasite in 1903. The bouts of fever of the early stage of the disease were associated with parasitaemic waves of trypanosomes in the circulation, while the enlarged lymph glands (notably those of the neck—Winterbottom’s sign), showed infiltration of parasites. The onset of mental symptoms in the late stage of the disease was associated with parasite invasion of the CNS and subsequent cellular infiltration. Further than this, the actual mechanisms by which trypanosomes cause disease remained obscure.

Gambian sleeping sickness originally appears to have been a widespread endemic disease in west Africa, but with the opening up of Africa to trade and colonization at the turn of the twentieth century came a series of devastating epidemics, notably in the Congo River basin and Uganda. Rhodesian sleeping sickness later appeared to spread in a series of focal epidemics throughout East Africa, possibly associated with movements of troops, refugees and migrant workers (Fig. 42.1) (Mulligan 1970).

From its discovery, Rhodesian sleeping sickness was always considered to be a zoonotic infection, since it could be contracted in areas inhabited only by wild animals (Ashcroft 1959). The proof of this came in 1958 when a human volunteer was infected with trypanosomes from a bushbuck in East Africa (Heisch et al. 1958). Subsequently, various wild and domestic animals were incriminated as reservoir hosts of the disease by in vitro methods (Onyango et al. 1966; Rickman and Robson 1970; Geigy et al. 1975; Brun and Jenni 1987) (Fig. 42.2). By contrast, Gambian sleeping sickness was held to be a disease of man alone, until evidence for the existence of animal reservoirs (pigs, dogs and antelope) was obtained in West Africa (Mehlitz et al. 1982).

Initially the Gambian and Rhodesian forms of sleeping sickness were thought to be associated with different vector species: the Gambian disease was believed to be transmitted only by tsetse flies of the palpalis group and the Rhodesian disease by flies of the morsitans group (Buxton 1955). However, palpalis group flies were found to be vectors of T. b. rhodesiense in outbreaks of sleeping sickness in Kenya and Ethiopia in the 1960s and 1970s, and it is now accepted that either or both fly groups may transmit each species.

Before the cause of sleeping sickness was discovered, control measures involved the natural movement of populations away from tsetse-infested zones as unhealthy settlements were abandoned, or isolation measures, such as quarantine and restricted movement, enforced by various colonial governments. Once the tsetse fly was realised to be the transmitting agent, evacuation of tsetse-infested areas was instituted. This was a practical measure for control of Gambian sleeping sickness, where the vector species were palpalis-group tsetse occupying limited areas of riverine or lakeshore vegetation. The first trypanocidal drugs became available soon after the turn of the twentieth century thanks to Ehrlich’s work (Atoxyl 1905; Suramin 1920) and drugs effective against parasites in the CNS followed in the 1940s (Melarsen 1940; Mel B 1947), allowing effective programmes of surveillance and treatment (Mulligan 1970).

Early control measures against morsitans-group tsetse involved bush clearance to destroy their habitat and game eradication to remove their food source. These measures were employed more widely to control animal trypanosomosis than the human disease. Bush clearance was effective but labour intensive and required continuous maintenance of cleared areas by settlement and farming to prevent reinvasion. Game eradication was first tried in a huge natural experiment brought about by rinderpest, which swept Southern and Eastern Africa in 1896. This dramatically reduced tsetse numbers by destroying large numbers of their favourite wild hosts. Later shoot out campaigns failed to remove small, more secretive animals such as bushbuck and bushpig, which by themselves are capable of maintaining large tsetse populations, and the strategy lost favour as concern for wildlife conservation grew (Ford 1971).

Insecticides came into widespread use in the 1950s to control both palpalis- and morsitans-group flies. Residual insecticides could be applied locally by knapsack sprayers to tsetse resting sites on vegetation fringing lakeshores and rivers or surrounding settlements. Widescale coverage from the air was only suitable for conditions of flat terrain and light vegetation, besides being costly. Again, environmental concerns now discourage large-scale insecticide use.

Kingdom: Protista; phylum: Sarcomastigophora; order: Kinetoplastida; family: Trypanosomatidae; genus: Trypanosoma; subgenus: Trypanozoon; species: T. evansi, T. brucei; subspecies: T. b. gambiense, T. b. rhodesiense, T. b. brucei. NB. This is the taxonomy in current widespread use, and, although the higher level taxonomy of protists is under revision, kinetoplastids and trypanosomatids, as well as genus Trypanosoma, remain as recognized groups (Cavalier-Smith 1981; Moreira et al. 2004).

All three subspecies of T. brucei are indistinguishable both by morphology and their lifecycle in the fly. All infect a wide range of mammalian hosts, including wild and domestic animals, but only T. b. rhodesiense and T. b. gambiense are capable of infecting man. The present subspecies all originally had species status, which can be confusing in older literature. The relationship between the three subspecies has always engendered heated debate. Some workers have proposed that T. b. gambiense and T. b. rhodesiense represent the same parasite under different epidemiological conditions, while others have argued that T. b. rhodesiense and T. b. brucei are one species. Long term transmission experiments demonstrated that T. b. rhodesiense retained infectivity to man despite extended sojourn in other animal hosts, while the fact that T. b. brucei had a far wider distribution than sleeping sickness indicated its separate identity. Extensive biochemical characterization data and the demonstration of genetic exchange now argue that T. b. rhodesiense and T. b. brucei represent a genetic continuum in East Africa, differentiated only by the trait of human-infectivity. This trait is now known to be conferred by a single gene—the serum resistance associated (SRA) gene—in T. b. rhodesiense (De Greef et al. 1989; Xong et al. 1998), which has become a convenient genetic marker for the subspecies. T. b. gambiense does not have the SRA gene and, although closely related to West African strains of T. b. brucei, has distinct biochemical and biological features, which make this homogeneous group of strains easily recognizable throughout its range.

Trypanosoma evansi is hard to distinguish from T. brucei ssp. morphologically, but lacks short-stumpy forms (Fig. 42.3) and is unable to develop in the tsetse fly. The kinetoplast DNA, though, is fundamentally different in the two species, as T. evansi lacks maxi-circles and has homogeneous mini-circles—a feature that has been exploited for the development of molecular markers. Although T. evansi does not have the SRA gene, it nevertheless may infect humans lacking ‘trypanolytic factor (TLF)’ in their serum, as demonstrated by a recent case of T. evansi infection in an Indian farmer (Vanhollebeke et al. 2006). This individual lacked apolipoprotein L-1, previously identified as TLF (Vanhamme et al. 2003), and it seems likely that this genetic trait would also confer susceptibility to infection with T. b. brucei and other tsetse-transmitted trypanosomes.

Trypanosoma b. brucei has become a favourite laboratory organism for molecular biologists: not only is it easy to grow and purify, but it has a number of biochemical peculiarities, some of which challenge textbook orthodoxy (Opperdoes 1985; Simpson 1990). With publication of the genome sequence of T. b. brucei in 2005 (Berriman et al. 2005), and those of the other subspecies to follow, T. brucei has become an important model organism.

Kinetoplastids are named after the densely-staining kinetoplast, a unique organelle containing the mitochondrial DNA. Kinetoplast DNA consists of a network of interlinked DNA circles of two sizes: about 5,000 mini-circles and 50 larger maxi-circles, carrying genes for mitochondrial function. A surprising finding was that maxi-circle transcripts are edited, rather than being faithfully transcribed as dogma would dictate. This clarified the role of the mini-circles, now known to encode some of the short RNAs which guide the editing process (Stuart et al. 2005), and helped explain the close association of mini- and maxi-circles; the benefit of RNA editing to the trypanosome remains obscure, however. Kinetoplast DNA is one target for trypanocidal drugs such as isometamidium and homidium which bind to DNA, possibly interfering with its replication or resulting in damage by double-strand breakage.

The surface of trypanosomes in the mammalian host is covered by a dense coat consisting of a single protein, the variant surface glycoprotein (VSG) (Borst and Cross 1982; Pays et al. 2004). The VSG is highly antigenic, but by changing its protein coat, the trypanosome can evade the immune response of the host, and this leads to characteristic waves of parasitaemia as antibody production catches up with each succeeding antigenic variant. Since an individual trypanosome has a repertoire of an estimated 1,000 genes coding for different VSGs, which also have a high mutation rate, vaccination is not a practical possibility. The trypanosome only expresses one VSG gene at a time in an expression site located at the end of a chromosome (telomere), which fits into a special compartment of the nucleus for transcription by RNA polymerase I (Navarro and Gull 2001). The mechanism of antigenic variation is complex, but in essence, antigenic switching is brought about when the active gene is displaced from the expression site by recombination with part or all of another silent VSG gene. Study of the transcription of VSG genes led to the discovery of discontinuous transcription in trypanosomes, a mechanism whereby a short (39 nucleotide) RNA leader sequence is spliced onto the 5’ end of all messenger RNAs.

Trypanosome chromosomes do not condense during the cell cycle, but intact chromosomal DNA molecules can be separated by pulsed field gel electrophoresis. The diploid karyotype of T. b. brucei consists of at least 100 chromosomes ranging in size from 50kb to several Mb (Melville et al. 1998). VSG genes together with housekeeping genes are found on chromosomes of all sizes, but only VSG genes are found on the smaller chromosomes. In particular, the 100 or so mini-chromosomes—small linear chromosomes of 50 to 150kb—constitute a reservoir of telomeric VSG genes, which may serve to promote rapid evolution of the antigenic repertoire.

Two other molecular features of trypanosomes may also provide new therapeutic avenues. Kinetoplastids are unique among eukaryotes in that glycolysis takes place inside a specialized organelle, the glycosome, which also contains enzymes for (parts of) other metabolic pathways. Gene sequencing has demonstrated that the glycosomal glycolytic enzymes are significantly different from their mammalian counterparts. Glutathione is an essential metabolite of eukaryote cells, which is reduced by glutathione reductase. In trypanosomes this enzyme was found to have a unique cofactor, trypanothione, a novel glutathione-spermidine conjugate. Inhibition of trypanothione synthesis or its dependent oxidation-reduction system would be expected to upset glutathione metabolism. This knowledge has already led to rational attempts at combination chemotherapy using existing drugs. Elfornithine (DFMO) inhibits ornithine decarboxylase, an enzyme in the polyamine biosynthetic pathway leading to trypanothione. The action of elfornithine can be augmented by combination chemotherapy with other drugs that interfere with this pathway, e.g. diminazene aceturate (Berenil). In the short term, this approach may well dramatically improve the efficacy of existing drugs, with the benefit of reduced dosage and hence toxicity, while the long term search for new trypanocidal compounds continues.

Trypanosoma brucei ssp. trypanosomes do not just circulate in the bloodstream, but invade various tissues most importantly the CNS, where they can more effectively evade the immune system. Trypanosomes interact in various ways with cells of the immune system, the complexities of which have yet to be fully unravelled, but are key to understanding the nature and extent of pathology and thus the course of disease (Rhind and Shek 1999; Vincendeau et al. 1999; Mansfield 2001). Trypanosomes multiply initially at the site of the fly bite and local inflammation may result in the first sign of the disease, the chancre. From here the trypanosomes invade the bloodstream and lymphatic system, causing generalized febrile attacks. Each parasitaemic wave is associated with expression of a limited number of VSGs; trypanosomes expressing these antigens are cleared from the blood by a specific antibody response, releasing further internal antigens. Subsequent parasitaemic waves are initiated by residual trypanosomes that have switched to expression of new surface antigens (VSGs). The sequential parasitaemic waves stimulate polyclonal B cell proliferation, particularly IgM, followed by T cell-dependent production of specific antibodies, which lead to destruction of parasites by the usual routes (complement-mediated lysis, phagocytosis, cell-mediated cytotoxicity). This initial robust immune response is followed in a chronic infection by exhaustion of the immune system and immunosuppression, an effect observed on both T and B cell responses. Macrophages are thought to play a major role in destroying trypanosomes in the tissues and are activated during infection by exposure to IFN-γ produced by CD8+ T cells, and trypanosome components, notably VSG. The activated macrophages secrete many components such as nitric oxide and reactive oxygen species capable of destroying trypanosomes, and various cytokines, notably TNFα, resulting in tissue damage and systemic effects. In animal models, the catabolic effects of TNFα (cachectin) give rise to anorexia, fever and weight loss, suggesting that it alone could be responsible for much of the clinical picture of trypanosomosis.

Invasion of the CNS commences at an early stage of infection and leads to progressive damage, culminating in coma and eventual death, if the disease remains untreated. In experimental animals dye-tracer studies show that the blood-CSF barrier is compromised within a few days of infection. An early change is the breakdown of the choroid plexus allowing infiltration of trypanosomes and lymphocytes into the circumventricular regions. No toxins are known to be released from the trypanosomes themselves; rather it appears to be the interactions of the parasites with the immune system that manifest disease. In particular, the cytokine/prostaglandin network appears to play a key role in the inflammatory processes in the brain, and also may be influencing wider physiological abnormalities such as somnolence and fever. Unravelling this complex series of interactions will eventually help to clarify late stage pathogenesis.

Trypanosomes are obligate parasites, which have no survival stages outside their hosts. In the mammalian host, trypanosome metabolism is dependent on high levels of glucose, as found in the bloodstream and CNS, while in the fly the main energy source is proline. Trypanosomes can survive for a limited time (hours) in samples of body fluids, e.g. blood, CSF, but are easily destroyed by heat, desiccation, detergents and disinfectants. There is no record of aerosol transmission.

Besides man, T. brucei ssp. can infect a wide variety of domestic and wild mammal species; natural infection has also been recorded in a bird (domestic hen) and a large reptile, the monitor lizard, which frequently serves as a source of blood meals for some tsetse species. In general T. b. rhodesiense and T. b. brucei are more virulent than T. b. gambiense and produce more rapid and severe symptoms. Animal trypanosomosis is also caused by other tsetse-transmitted trypanosome species in Africa, notably T. vivax, T. congolense and T. simiae, which are considered to be of far greater veterinary importance than T. brucei ssp., except perhaps for dogs, horses and camels (Stephen 1986; Brown et al. 1990). Mixed infections of two or more trypanosome species occur frequently in livestock kept in endemic zones. Outside sub-Saharan Africa, Trypanosoma evansi assumes great importance as a pathogen of livestock, particularly camels and horses, but also cattle and buffalo, and has occasionally caused the death of zoo carnivores fed on infected carcasses; T. evansi is included here, because of a single, well-documented case of human infection in an Indian farmer (Joshi et al. 2005).

The incubation period in both man and animals is highly variable due to differences in trypanosome virulence and number of organisms inoculated, besides individual and species differences in susceptibility and previous exposure to tsetse challenge. It is also difficult to assess in an endemic area where fly bite is frequent. For the most virulent strains in susceptible hosts, the incubation period is probably as little as a week, but may be several weeks or much longer; prolonged incubation periods of several years are on record for Gambian sleeping sickness. Trypanosome infection in animals may remain cryptic for years, as evidenced by the development of disease in zoo animals several years after removal from a tsetse-endemic zone.

Generally the early symptoms of Rhodesian sleeping sickness tend to be more severe and acute than those of the Gambian form, and the early and late stages are less clearly demarcated (WHO 1998). The early stage is characterized by intermittent fever, weakness, headache, backache, joint pains, oedema, pruritis, and enlargement of the lymph glands and spleen, while the late stage is marked by neurological symptoms and endocrine disorders, e.g. amenorrhea or impotence. However, due to the early invasion of the CNS, neurological signs, such as facial ticks and mood and appetite changes, may be present at an early stage. Two of the first signs of sleeping sickness are the chancre, an indurated swelling at the site of fly bite which may be present in early cases of Rhodesian sleeping sickness, but is seldom observed for Gambian sleeping sickness, and swelling of the lymph nodes, particularly the cervical glands ‘Winterbottom’s sign’. The eponymous sleep disorders include nocturnal insomnia and daytime somnolence, and classically sleeping sickness manifests itself finally in coma. Cerebral involvement may also be evident from psychiatric disturbances, ranging in severity from behavioural changes, often increased aggression or violence, to frank psychosis. Late-stage patients have ended up in mental institutions or even prison for these reasons.

Some animals infected with T. brucei ssp. (e.g. horses, dogs) may be obviously ill, but others, particularly bovids and wild animals, may remain asymptomatic, possibly showing disease only if stressed (Brown et al. 1990). There is the added complication that when a mixture of trypanosome species is present, the presence of T. brucei ssp. may be easily missed. Early infection is characterized by intermittent parasitaemia associated with bouts of fever. Clinical features of the chronic disease include anaemia, fever, cachexia, lymphadenopathy, and oedema. Nagana was the Zulu name for animal trypanosomosis and means ‘a state of depressed spirits’, which nicely sums up the clinical picture. Lameness of the hindquarters is characteristic of infection in dogs and horses, as is corneal opacity—so-called ‘white eyes’—in dogs. There may also be disorders of reproduction, particularly sterility and abortion. A similar clinical picture is observed in camels, horses and dogs with Surra, while T. evansi infection may remain cryptic in less susceptible animals such as bovids.

For human patients parasite demonstration is generally required before treatment commences as drug treatment is not without risk. For animal trypanosomosis, this is not always necessary, and for example, herd treatment may be carried out after demonstration of trypanosomes in some individuals only. Much the same diagnostic methods are used for both human and animal trypanosomosis, but levels of parasitaemia may often be very low and fluctuating. Parasitaemias tend to be higher in Rhodesian sleeping sickness and therefore it is usually possible to find motile trypanosomes by simple microscopic examination of wet blood films. A trypanosome concentration method is usually necessary for demonstration of T. b. gambiense, the simplest being thick blood film stained with Giemsa or Field’s stain; other concentration methods include mini-anion exchange columns, haematocrit (HTC) buffy coat and QBC (quantitative buffy coat technique); for animal use, HTC has the advantage that packed cell volume can be measured at the same time to provide an assessment of the degree of anaemia. If enlarged lymph glands are present, trypanosomes can often be demonstrated by gland puncture; this method has also been used for cattle. However, trypanosome numbers may be extremely low and the most reliable and sensitive means of demonstrating parasites is then by inoculation of rodents, immunosuppressed in the case of T. b. gambiense.

Serological tests such as the CATT (Card Agglutination Test for Trypanosomosis) (Magnus et al. 1978), which rely on antibody detection, are useful for preliminary screening in endemic areas of Gambian sleeping sickness, but suspects require parasitological confirmation before treatment. New diagnostic methods are constantly being sought, and an ELISA test for antigen detection, various PCR (polymerase chain reaction)-based tests and LAMP (loop-mediated isothermal amplification) tests have all been developed in recent years. The cost and ease of use of any new diagnostic test in the field are factors always to be borne in mind.

In man, CNS involvement is assessed by microscopic examination of centrifuged CSF withdrawn by lumbar puncture. If no parasites can be demonstrated, then high numbers of cells (more than 5 per ml) or high levels of protein (greater than 37mg/100ml) indicate CNS involvement (WHO 1998). In exceptional cases, where trypanosomes can neither be demonstrated in the blood or CSF, diagnosis may be made on clinical grounds only.

Much of the recent work has been carried out in animal models, which appear to share similar pathology with man; the older post-mortem data on patients who died in late stage disease was reviewed by Kristensson and Bentivoglio (1999). Trypanosoma brucei ssp. trypanosomes invade the intercellular fluids of various tissues, as well as the bloodstream and extracellular fluids. In the early haemolymphatic stage of sleeping sickness, the lymph nodes and spleen are enlarged and infiltrated with lymphocytes, plasma cells and monocytes (Greenwood and Whittle 1980; WHO 1998). Later the lymph nodes become shrunken and atrophied with progressive exhaustion of the immune system. Anaemia is haemolytic and large numbers of reticulocytes are present. There is also thrombocytopaenia. The intercellular spaces of various tissues are invaded, notably the heart, with resultant myocarditis and pericardial effusions. Invasion of the CNS begins with congestion of the meninges and infiltration of lymphocytes and large vesiculated cells. These are the so-called morular cells originally observed by Mott in 1905, which are plasma cells containing huge amounts of immunoglobulin. Inflammation extends into the brain tissue and blood vessels show perivascular cuffing. There is proliferation of neuroglial cells (astrocytes and microglia) associated with diffuse meningoencephalitis.

Unfortunately, all the drugs currently used for treatment of human sleeping sickness are rather toxic and cause side effects ranging from unpleasant to severe; only one new drug—eflornithine—has been introduced in the past 60 years (Pepin 2007), although several are now in the pipeline. Early cases without CNS involvement are treated with suramin or, for Gambian sleeping sickness only, pentamidine (Lomidine). Some use has also been made of diminazene aceturate (Berenil), although this drug is not registered for human use. None of these drugs cross the blood-brain barrier to any extent and late stage cases require one or more courses of the arsenical melarsoprol (Arsobal). This drug is dissolved in propylene glycol and leakage into the tissues at the injection site causes severe irritation. Relapses following melarsoprol treatment are problematic as the patient may already be in a poor state; relapsed cases of Gambian sleeping sickness can be treated with elfornithine (DFMO; Ornidyl), but this drug is not effective against T. b. rhodesiense. Nifurtimox in combination with melarsoprol has produced encouraging results for treatment of late stage Gambian sleeping sickness (Bisser et al. 2007).

For animal trypanosomosis, three drugs are in common use for treatment of ruminants—Berenil, isometamidium (Samorin, Trypamidium) and homidium salts (Ethidium, Novidium). Of these only Berenil and Samorin are recommended for treatment of T. brucei ssp. infections. For horses and camels, Samorin, suramin (Naganol) or quinapyramine sulphate (Antrycide, Trypacide) are the drugs of choice and Samorin is recommended for treatment of dogs. Samorin and Trypacide Prosalt (quinapyramine sulphate together with the more insoluble chloride salt) are recommended for prevention as well as cure in cattle and horses, the effects lasting 3–6 months. There is some degree of drug resistance to all veterinary trypanocidal drugs and the small number available make this possibility a constant concern.

Untreated sleeping sickness is fatal, with death resulting in 3–9 months with Rhodesian sleeping sickness and possibly a matter of years with Gambian sleeping sickness. In cases without CNS involvement, prognosis is generally good; however, such cases may relapse if CNS involvement was unrecognized at the time of treatment. The state of some patients admitted with late stage disease may already be so poor that they require general nursing and supportive therapy before commencement of treatment. The most severe complication of treatment in late stage patients is so-called reactive encephalopathy, which occurs in 5–10% of cases and leads to high mortality. In the past, this syndrome was considered to be a severe side effect of arsenical treatment. However, it also occurs when other drugs are used for treatment of late stage disease. This could indicate that it is a reaction to the rapid release of trypanosomal antigens in the CNS as the trypanosomes are killed, but work in an experimental mouse model suggests the cause to be the persistence of small numbers of live trypanosomes in the brain. The latter would indicate the beneficial effect of aggressive rather than gradual drug therapy. The results of reactive encephalopathy can be ameliorated by supportive treatment with anti-inflammatory drugs.

All patients need to be followed up after treatment for at least a year to ascertain whether cure has taken place. WHO (1998) recommend follow up examinations at six monthly intervals for two years. A full parasitological and clinical examination is necessary, including lumbar puncture.

In animals the course and outcome of T. brucei ssp. infection is highly variable, depending on the subspecies and strain of infecting trypanosomes, the species and breed of mammalian host and its previous exposure to trypanosomes. In horses and dogs the disease may take an acute and fatal course within a few weeks. Severe chronic trypanosomosis leads to progressive debility and emaciation and death in a matter of months. Alternatively, infection may be transient. Drug therapy is generally curative, but relapses may arise from residual parasites hidden in the tissues or from drug resistance.

The number of new cases of sleeping sickness is estimated to be in the tens of thousands annually, but accurate figures are difficult to obtain as many endemic foci are located in remote and inaccessible areas, with poor health facilities (WHO 2008). It is costly to maintain surveillance in such regions, but control programmes may also fail for other reasons such as civil disturbance, war, or other economic and political problems.

Accurate data would need to be gathered over a period of years to gain a true picture of the prevalence of sleeping sickness in an endemic focus. As noted above, such data are hard to come by for humans, and even harder to obtain for potential animal reservoir hosts. The activity and distribution of tsetse flies varies according to wet or dry seasons, with consequent effects on transmission rates and presumably prevalence. There may also be seasonal movements of domestic stock or wildlife. On top of this, survey work may be restricted, if not impossible, during the wet season in remote areas.

Sampling from animal species present in a focus is often unrepresentative, depending on the ease with which they can be caught. Conservation measures in some countries mean that a licence is required to sample wild animals. Some information on prevalence can be obtained from tsetse flies caught in an endemic area by examining trypanosome infection rates, together with blood meal identification.

Most endemic foci smoulder on with a low annual incidence and occasional flare ups (Kuzoe 1993). The usual reason for recrudescence of foci is breakdown of routine control measures. Sometimes control programmes are discontinued because the economic costs of control appear to outweigh the benefits of dealing with relatively few cases or small numbers of flies. This can be a false economy due to the high cost of controlling any ensuing epidemic. Some outbreaks result from prolonged civil disruption, when routine control measures break down and aid agencies withdraw financial and technical resources. Reliable statistics are then hard to come by. In these circumstances, people may flee their homes and seek refuge in areas where they are more at risk of tsetse bite and less likely to receive medical help. Recent epidemics in the Democratic Republic of Congo (DRC), Angola and Uganda reflect past civil strife. Government resettlement schemes or unofficial settlement adjacent to designated wildlife areas have also resulted in new outbreaks.

There are considered to be 60 million at risk living in the 250 or so endemic foci of sleeping sickness scattered throughout sub-Saharan Africa (Fig. 42.1). Occupational risk groups include hunters, firewood or honey gatherers and tourists to wildlife parks in endemic areas in East Africa, who may contract sporadic Rhodesian sleeping sickness from wild animal reservoirs (Jelinek et al. 2002). No specific occupational groups, except perhaps fishermen, are at risk for Gambian sleeping sickness, since all sectors of the population come into contact with the fly in its riverine/lakeshore habitat during daily activities such as bathing and collecting water.

Population movements may also increase risk (Ford 1971). For example, resettlement schemes may bring people into close contact with flies which previously fed on wild animal reservoir hosts. Refugees fleeing from one country may move into an endemic focus in another, as happened in the 1990s on the Sudan-Uganda-DRC border spanned by an endemic focus of Gambian sleeping sickness. Immigrants are at no greater risk of infection than the indigenous population, since exposure does not confer immunity to re-infection. Returning refugees may find their abandoned farmland overgrown with bush and infested with tsetse, giving rise to new outbreaks.

Sleeping sickness has been reported in 36 countries in sub-Saharan Africa between latitudes 14^oN and 29^oS (WHO 2008). The distribution of the disease is restricted by that of the tsetse fly, which needs the right conditions of humidity and temperature for survival (Leak 1999). Palpalis-group flies, which are the main vectors of Gambian sleeping sickness, need high humidity and are found in the forested zones of West and Central Africa, Uganda and southern Sudan; their range also extends northwards through more arid country, following the lines of fringing vegetation on river and lakeshore. Morsitans-group flies can tolerate drier conditions and are found throughout the wooded savannah regions.

The source of infection is always another infected mammalian host. In an epidemic, other infected humans are generally thought to be the major source of infection, although it is acknowledged that a high infection rate in domestic livestock could boost transmission, if flies take a significant proportion of feeds from both hosts. At the other end of the spectrum, sporadic infections can be contracted

when man breaks into a wild animal-tsetse cycle. This occurs most frequently in East Africa, where there are large concentrations of wild animals in uninhabited bush, but this possibility cannot be ruled out in some areas of West and Central Africa with plentiful wildlife. Domestic livestock will probably become increasingly important in disease transmission, either acting as an intermediary in transferring trypanosome strains from wild animals to man, or by replacing wild animal reservoirs altogether. The single human case of T. evansi infection in an Indian farmer was probably contracted from infected livestock (cattle) (Joshi et al. 2005).

The usual mode of transmission is via the bite of an infected tsetse fly (Fig. 42.4), as can be deduced from the restriction of the disease to the area of tsetse infestation. Trypanosomes undergo a developmental and multiplicative cycle in the fly, first in the midgut and then in the salivary glands, from whence they are conveyed to new hosts with the saliva (Hoare 1972). Flies are relatively refractory to infection, only readily becoming infected at the first blood meal after emergence from the puparium (Maudlin 1991); consequently much less than 1% of flies are found infected with T. brucei ssp. in the wild. The cycle takes a minimum of about two weeks to complete. With a lifespan of two to three months under favourable conditions and a requirement to feed every few days, an individual fly could infect at least 20 new hosts. It is for this reason that close man-fly contact, rather than sheer numbers of flies, is so important in the epidemiology of sleeping sickness: the classic example is of one resident infected fly at the village water hole with the potential to cause a small epidemic in the dry season when people spend more time there (Nash 1969). By contrast, large concentrations of flies feeding predominantly on wild animals may pose very little risk.

Other modes of transmission are possible, but depend on relatively high parasitaemia in the donor and a susceptible recipient, e.g. contaminative transmission by tsetse or other biting flies such as tabanids (= mode of transmission of T. evansi); direct transmission via fresh infected blood or raw meat (Moloo et al. 1973). Congenital transmission has rarely been recorded.

The main route of sleeping sickness transmission is via the tsetse fly. Laboratory studies have shown not only that different trypanosome subspecies and strains vary in infectivity to flies, but also that the flies themselves differ in susceptibility to trypanosome infection (Maudlin 1991). In fact it appears that most flies are refractory to infection and transmission appears to be sustained by relatively few flies. The long-term stability of endemic foci is evidence that this minimalist strategy is successful and further emphasizes the fundamental importance of the behaviour of individual flies rather than populations in transmission of the disease.

Establishment of infection in a new host following fly bite will depend on several factors: the number of trypanosomes inoculated with the saliva; the virulence of the trypanosome subspecies or strain; the level of intrinsic resistance of the host.

The risk of contracting human trypanosomosis outside the known foci of the disease is minimal, although sporadic infections could potentially be acquired in some tsetse-infested wildlife areas. Animal trypanosomosis exists throughout the tsetse belt, however, and T. evansi is a widespread pathogen of livestock in parts of N. Africa, the Middle East, Asia and S. America; only in very rare cases have these animal trypanosomes ever been isolated from humans.

Even in an endemic area, the risk of being bitten by an infected fly is small. The risk of infection can be lowered by avoiding tsetse fly bite, easier said than done in the African bush. All tsetse species require some degree of shade and habitats vary from light bush to dense forest and even conifer or coffee plantations. Tsetse can be found in the peridomestic environment if there is suitable cover and those species favouring a riverine habitat may also be encountered at waterholes, river crossing points and bathing places. Tsetse feed during the day and find their hosts by sight as well as smell. They are attracted to large moving objects, particularly vehicles, and also dark colour clothing. Clothing is no barrier to tsetse bite, but insect repellents might be of limited use if practical. Animals can be treated with pour-on residual insecticide, but this might not prevent fly bite and the possibility of infection.

Prophylactic drugs for human use cannot be recommended, although pentamidine was widely used in the past (Mulligan 1970). There would be too great a risk of masking early infection, giving the parasites time to invade the CNS. For animals, various drugs can be given prophylactically, but there could be toxic effects from long-term use and prophylaxis would probably be ineffectual in areas of high challenge.

There are two current strategies for control: control of the parasite by regular medical surveillance of the population at risk or control of the fly. The parasite is targeted by drug therapy of its human host; there have been few if any attempts to eliminate trypanosomes in reservoir hosts, other than early attempts to destroy the wild animal reservoir. Eradication of tsetse flies in Africa remains a distant hope. Despite their slow breeding cycle—each female produces only a single larva every 9 days or so—and the various measures used to combat them, tsetse fly populations remain buoyant and tsetse have actually increased in distribution in some areas. Present environmental concerns favour limited use of insecticides and low-tech methodologies that can be widely and cheaply applied at local level; for example, the use of locally-made traps or targets (Dransfield et al. 1991). On the larger scale, Sterile Insect Technique (SIT) was used successfully to eradicate tsetse from Zanzibar island in a four year campaign started in 1994; whether tsetse control by SIT will prove either possible or economic on the African mainland is hotly debated (Rogers and Randolph 2002).

Sleeping sickness control is usually organized at governmental level; however other organizations, such as aid agencies and missions, may be involved, especially during epidemics. For example, besides local government, several European governments, UN organizations and charities have all provided financial or technical aid for control of recent epidemics in south east Uganda, DRC and Angola. International organizations such as Organisation of African Unity (OAU) and WHO take responsibility for coordination of control measures between African countries. This is especially important where foci span borders and for communication between French- and English-speaking countries.

Case detection and treatment is possible at a number of organizational levels, with various degrees of cost-effectiveness. At one extreme, mobile teams can be used to survey endemic areas screening the whole population every year. Active surveillance like this is very costly, but also fulfils the requirement for patient follow-up to assess cure. Such programmes are necessary to cover large areas of endemic Gambian sleeping sickness. At the other extreme, patients can be left to report to their local health centre. This is so-called passive surveillance and is adequate for small outbreaks of Rhodesian sleeping sickness.

Individual governments must decide the level of resources to devote to sleeping sickness surveillance on the basis of the population at risk, the level of endemicity and their own health priorities. Present pressures on health budgets and the relatively small number of cases mean that sleeping sickness is often seen as a low priority. However, since it is simpler and less expensive to undertake regular surveillance than deal with an explosive epidemic, the risk of epidemics makes sleeping sickness a major public health problem in Africa. Epidemics cause fear in local populations, since the etiology of the disease is still mysterious, and have both social and economic consequences, such as depopulation and abandonment of farmland.

Methods for controlling tsetse flies currently comprise insecticide application by knapsack sprayers or by air, trap/target technology, bush clearance, or sterile male release (SIT). In planning an insecticide campaign, both adult flies and offspring must be considered. Females deposit their fully grown larvae on the ground in suitable sites where the soil is loose and moist; this is to facilitate rapid burrowing of the larva into the ground, where it pupates (Buxton 1955). Survival of pupae depends greatly on humidity and temperature and adult flies emerge roughly 3 weeks later. Thus, a single application of non-residual insecticide will have little effect on the tsetse population. Aerial spraying of insecticide (e.g. endosulphan) necessitates several spray cycles at weekly intervals. Effective knapsack spraying uses residual insecticides (e.g. dieldrin) applied to tsetse resting sites on foliage and branches. In the dry season tsetse populations retreat to the most favourable parts of their range making control far more cost-effective. Thorough knowledge of the distribution and habits of the tsetse species to be controlled is invaluable.

Environmental concerns have put widespread insecticide use into disfavour and promoted the development of trap and target technology (Fig. 42.5). Traps were originally used as a sampling method for tsetse populations and work on the principle that tsetse are attracted by sight to large dark objects. Whether this is in order to seek shade or a host is unclear, as one optimum but curious design is a biconical trap made of dark blue cloth. The tsetse enter at the bottom and move upwards through the trap towards the light, where they are imprisoned in a small cage and soon succumb to death by desiccation or hunger. Targets consisting of simple square sheets of the same colour cloth are also attractive and can be impregnated with insecticide (e.g. deltamethrin) to kill the flies as they land. Odour attractants (constituents of cow breath or urine) can be used in combination with traps or targets to increase the catch. These methodologies, although environmentally sound and low cost, are labour intensive. Large numbers of traps/targets need to be deployed to give adequate coverage and they require regular repair and replenishment of odour baits or insecticide. For these reasons, trap/target methodology is well suited to programmes involving community participation (Dransfield et al. 1991). Ideally traps or targets are constructed locally out of indigenous or cheap materials and are maintained by the community.

Bush clearance permanently destroys the habitat of tsetse and is cost effective when new farmland is gained. Sometimes, though, farming practices actually create new tsetse habitats: for example, the widespread introduction of Lantana hedges or plantations of conifers, coffee or cocoa.

Control by sterile male release or SIT depends on the fact that females mate once only and store the sperm until required. Sufficiently large numbers of sterile males in competition with wild males will thus reduce the population density over time. This technique is essentially long-term and requires high capital input initially to set up facilities for rearing and irradiating large numbers of male tsetse. The area of release needs to be isolated by geographical features or bush clearance in order to prevent re-invasion, and other techniques such as trapping or insecticides may need to be used in conjunction to push the tsetse population below recoverable levels. Hence, SIT was ideally suited to eradication of the single tsetse species on the island of Zanzibar, too far from the Tanzanian mainland for re-invasion. Multiple tsetse species and real possibilities for re-invasion, as well as high levels of capital cost, present enormous challenges for application of SIT elsewhere on the African continent.

Control of tsetse is often of both medical and veterinary concern, and this can lead to problems of communication if more than one government department or agency is involved. Sustained control programmes will be more cost-effective than emergency action in the long run. For example, regular application of insecticide by teams with knapsack sprayers or schemes to involve rural communities in making and maintaining tsetse traps are long-term, low-cost methods of control; application of insecticide from the air can rapidly and dramatically reduce transmission, but is a short-term, costly measure, unless the spraying permanently eliminates tsetse from the area.

The success of control programmes should be measurable by a decreasing incidence of sleeping sickness and, indeed, numbers of sleeping sickness cases have steadily dropped over the past ten years from the very high levels recorded in the 1990s. However, this was in part due to emergency interventions by international organizations to control epidemic disease. Resurgence will be prevented in the long-term by regular survaillence and fly control, difficult to maintain in countries where health resources are strained to the limits and there are other health priorities.

In principle, community participation schemes should devolve responsibility for control activities to local level thereby rendering them sustainable. In practice, without continued technical input, communities lose interest in control schemes as the problem becomes less acute.

There is no legislation to control sleeping sickness in most endemic countries and cases are not adequately reported. Within tropical Africa it is potentially possible that movement of infected livestock between countries in the tsetse-infested zone could create new foci of disease, and disease spread has certainly happened in this way within Uganda (Hutchinson et al. 2003). The translocation of wildlife species from endemic to non-endemic areas is another potential risk. Infection with tsetse-transmitted trypanosomes is generally not considered to be a risk outside tropical Africa, since the chance of transmission is minimal. Cases are imported into Europe from time to time, usually in those who have visited wildlife areas.