Several methods have been described to identify drug resistance in trypanosomes (reviewed by Peregrine, 1994; 1996). At present, three types of technique are commonly used to identify drug resistance: tests in ruminants; tests in mice; and in vitro assays. None of these is, however, an ideal test and other tests are still in the phase of development or validation. The advantages and disadvantages of each of the different techniques are briefly summarized in the following sections.
Tests in ruminants provide direct information from studies in ruminants using recommended doses of trypanocide. The tests commonly consist of infecting a group of cattle or small ruminants with the isolate under investigation and later, when the animals are parasitaemic, treating them with various levels of trypanocide. The animals are then regularly monitored over a prolonged period (up to 100 days) to determine the effective dose (ED), i.e. the dose that clears the parasites from the circulation, and the curative dose (CD), i.e. the dose that provides a permanent cure (Sones, Njogu and Holmes, 1988). For these studies, the cattle or small ruminants must be kept in fly-proof accommodation or in a non-tsetse area in order to eliminate the risk of reinfection during the study. A variation of this technique was used by Ainanshe, Jennings and Holmes (1992) in Somalia to examine a group of isolates from a district. Blood from a group of infected cattle was inoculated into a single recipient calf, which was monitored, and later, when it was parasitaemic, treated with trypanocide at the recommended dose. A breakthrough infection, indicative that one of the inoculated trypanosome populations was drug-resistant, was inoculated into groups of calves and mice to determine the level of drug resistance. This technique is useful in situations where laboratory facilities are very limited but it only allows a qualitative assessment and does not indicate how many of the isolates inoculated into a single calf were resistant.
Further constraints to this technique are that not all populations might grow equally well and that sensitive isolates might overgrow resistant ones when inoculated together (Sones, Holmes and Urquhart, 1989). However this is not a consistent observation (Burudi et al., 1994).
A useful indication of the level of resistance can be obtained from studies in ruminants (and mice) by recording the length of time between treatment and the detection of breakthrough populations of trypanosomes. The shorter the period, the greater the level of resistance (Ainanshe, Jennings and Holmes, 1992; Williamson and Stephen, 1960).
The advantages of studies in ruminants are that most trypanosome isolates of cattle are able to grow in these hosts and that the data obtained are directly applicable to the field.
The disadvantages are the long duration (a follow-up of 100 days is necessary to allow the detection of relapses) and the cost (purchase and maintenance of the animals are expensive). Furthermore, if only one isolate per animal is tested, it is usually impractical and too expensive to examine a large number of isolates.
After expansion of an isolate in a donor mouse, groups of five or six mice are inoculated with trypanosomes. Twenty-four hours later, or at the first peak of parasitaemia, each group except the control group is treated with a range of drug doses. Thereafter, the mice should be monitored three times a week for 60 days.
The ED50 or ED95 (the effective dose that gives temporary clearance of the parasites in 50 or 95 percent of the animals, respectively) can be calculated, as can the CD50 or CD95 (the curative dose that gives complete cure in 50 or 95 percent of the animals, respectively). Sones, Njogu and Holmes (1988) used groups of five mice, which allowed an easy calculation of ED80 and CD80 values (one out of five mice was not cleared or cured). These figures should be compared with those obtained using reference-sensitive trypanosome strains.
The advantage of the mouse assay is that it is cheaper than the test in cattle. There are several disadvantages, however: i) most T. vivax isolates, and also some T. congolense isolates, do not grow in mice; ii) although there is reasonable correlation between drug sensitivity data in mice and in cattle, higher doses of drug must be used in mice (normally ten times higher) in order to obtain comparable results to those obtained in cattle because of the vast difference in metabolic size (Sones, Njogu and Holmes, 1988); iii) precise assessment of the degree of resistance needs a large number of mice per isolate, which makes it a labour-intensive test - identification of a discriminatory dose, above which an isolate should be considered as resistant, could drastically reduce the number of mice and the amount of work to be carried out; and iv) it takes as long as 60 days to evaluate the drug sensitivity of an isolate.
Since the review of Kaminsky and Brun (1993) further progress has been made in the field of in vitro assays to determine the drug sensitivity of trypanosomes. These authors advised the use of metacyclic or bloodstream forms instead of procyclic forms in such assays. However, it takes up to 40 to 50 days of in vitro incubation to generate metacyclic trypanosomes (Gray et al., 1993). The advantage of this technique is that large numbers of isolates can be examined; tests with metacyclic trypanosomes correlate well with field observations. However there are several disadvantages. In vitro cultivation of bloodstream forms is only possible using preadapted lines and not using isolates directly from naturally infected animals (Hirumi, Hirumi and Peregrine, 1993). A simplified axenic culture system has been developed by these authors, but further research is still necessary to study the correlation with field data. A potential problem associated with this lengthy time of adaptation is the possible selection against trypanosomes that have the phenotype of the original population. In vitro assays are expensive to perform and require good laboratory facilities and well- trained staff.
If better techniques can be developed in order to adapt isolates more rapidly to grow in vitro, these assays may become more popular, especially in those laboratories where culture facilities are already established.
As an alternative to the tests mentioned above, the use of trypanocidal drug enzyme-linked immunosorbent assays (ELISAs) in combination with parasite detection tests has given promising results for the detection of resistant trypanosomes. A competitive ELISA which allowed the detection of small amounts of isometamidium in serum of cattle was first described by Whitelaw et al. (1991). This technique was further improved by Eisler et al. (1993) and Eisler, Elliott and Holmes (1996) and has been validated in cattle under experimental and field conditions (Eisler, 1996; Eisler et al., 1994; 1996; 1997a). The test is both sensitive, detecting subnanogramme concentrations, and specific. It allows the monitoring of drug levels over extended periods and the evaluation of factors influencing drug disappearance rates from the plasma.
The available data indicate that there is a considerable individual variation after intramuscular injection of ISMM in cattle (Eisler, 1996). One interesting finding has been that the drug disappears more rapidly in animals challenged and becoming infected with drug-resistant trypanosome isolates than in those challenged but protected against infection with sensitive trypanosomes (Eisler et al., 1994). Observations showed that the presence of trypanosomes in animals with an ISMM concentration of > 0.4 ng/ml suggests resistance; the higher the drug level detected the greater the degree of resistance that could be inferred (Eisler et al., 1997a). Further research is necessary, however, in order to confirm these results in a larger number of animals. Similar drug ELISAs have been developed for the detection of subnanogramme amounts of homidium bromide (Murilla, 1996) and a similar test for diminazene is in development.
The advantage of the ISMM ELISA is that large numbers of sera can be tested within a relatively short time. The ELISA may also provide information on drug usage in an area of investigation. The disadvantage is that further studies are required to confirm the correlation between protection against tsetse challenge with various trypanosome populations and the ISMM concentration in the serum. It is not yet possible to draw firm conclusions on the sensitivity or resistance of a trypanosome population at the level of the individual animal. The ELISA should, however, give some indication of the resistance situation at the level of the herd. A further disadvantage is that, while the ELISA may indicate the level of drug withstood by a trypanosome population, it does not provide information about the level required for protection.
Longitudinal parasitological data can be used to detect resistance problems, although their use is not suitable as a routine test. Rowlands et al. (1993) showed that the application of a computer model to parasitological data collected over a long period on a monthly basis allowed the incidence of new infections to be distinguished from recurrent infections. This analysis showed that the prevalence of diminazene-resistant infections in the Ghibe valley, Ethiopia, increased from 6 percent in 1986 to 14 percent in 1989.
The advantage of these kinds of data is of course that they are directly applicable to the field. The disadvantages, however, are that: i) the true prevalence of drug-resistant infections seems to be underestimated; ii) it is retrospective by at least six months; and iii) the technique is quite expensive, if a longitudinal study is not carried out for other purposes (Peregrine, 1996).
Since it has been shown that the rate of ISMM accumulation in T. congolense is a good indicator of the degree of drug resistance and since the mitochondrial electrical potential (MEP) appears to be closely linked with the rate of drug uptake, it might be possible in the near future to develop a quantitative in vitro test to evaluate the MEP (ILRI, 1996; Wilkes et al., 1997). If such a test could be carried out using a small number of trypanosomes, it might provide a rapid indication of the level of resistance of a given trypanosome isolate. It is hoped that this test could be conducted on whole blood samples and would not, therefore, suffer from the same disadvantages as other in vitro tests referred to earlier. An alternative approach in the longer term may be made to identify genetic markers for ISMM resistance, which might be developed into reagents for the identification of resistant trypanosomes using polymerase chain reaction (PCR).
Unfortunately, it will take several years before such a test is validated and becomes available to potential users. Standardization of the existing tests should, therefore, receive high priority, especially of the assays in mice and in the definitive hosts, because these can be carried out in less well-equipped laboratories. Such standardization should allow the establishment of a resistance monitoring system capable of comparing data on a temporal and spatial basis across Africa and avoiding the current lack of reliable data in the field of drug resistance of trypanosomes.