Evidence for bz resistance in human parasites

Reduced efficacy of BZ anthelmintics against soil transmitted nematodes has been reported, but not so far confirmed, by biological tests and genomic analysis (Albonico et al. Reference Albonico, Engels and Savioli2004a). De Clercq et al. (Reference De Clercq, Sacko, Behnke, Gilbert, Dorny and Vercruysse1997) reported a failure of MBZ to eliminate human hookworm infection in Mali, but did not provide follow up confirmation that the failure was due to resistance. Similarly, Albonico et al. (Reference Albonico, Bickle, Ramsan, Montresor, Savioli and Taylor2003) reported inefficacy of MBZ against intestinal nematode infections in people in Zanzibar who had been repeatedly treated. ABZ resistance has been reported in the intestinal protozoa, Giardia (Upcroft et al. Reference Upcroft, Mitchell, Chen and Upcroft1996; Reynoldson et al. Reference Reynoldson, Behnke, Gracey, Horton, Spargo, Hopkins, Constantine, Gilbert, Stead, Hobbs and Thompson1998). These authors did not find that ABZ resistance was associated with a codon 200 SNP in β-tubulin, as is commonly seen in BZ resistance in trichostrongylid nematodes of livestock. However, in the human filarial nematode, Wuchereria bancrofti, the Phe200Tyr SNP has been found in β-tubulin and increased dramatically in microfilariae obtained from patients in Burkina Faso, after combination treatment with ABZ and IVM (Schwab et al. Reference Schwab, Boakye, Kyelem and Prichard2005). Nevertheless, parasitological confirmation of BZ resistance in W. bancrofti is still lacking, although the treated people still had microfilariae in their blood seven days after the chemotherapy.

The difficulty of analyzing for bz resistance in nematode parasites of people

It is difficult to confirm unequivocally that anthelmintic resistance occurs in human helminth parasites, particularly in filarial parasites, such as W. bancrofti, which have no free-living stages and cannot be cultured in model hosts. In veterinary medicine, unequivocal evidence of anthelmintic resistance has usually come from collection of the egg or larval stages in animal faeces, culture to the infective stage and experimental infection of parasite-naïve hosts with the infective stage, followed by anthelmintic treatment and post treatment necropsy of the experimentally infected and treated animals. Such strategies for confirmation of anthelmintic resistance are impossible for several obvious reasons with parasites of humans unless the human parasite can be cultured and used to infect model animal hosts. In general, this is not possible.

It is also difficult to undertake in vitro biological assays for anthelmintic susceptibility for many human helminth parasites. The nematodes that cause lymphatic filariasis (LF) have no free living stages to enable, for example, larval development assays to be performed. Nor can the parasite that is the principal cause of LF, W. bancrofti, be maintained in any experimental animal host in which anthelmintic response assays could be conducted. Therefore, parasitological evidence of resistance can only be obtained by following the response of this parasite to anthelmintics, in humans, over a long period of time. The most important effect of ABZ on W. bancrofti is the prolonged suppression of reproduction (for approximately 9 months) by the adult parasites (Michael et al. Reference Michael, Malecela-Lazaro, Simonsen, Pedersen, Barker, Kumar and Kazura2004). To assess shortening of the duration of suppression of reproduction, it would be necessary to determine blood microfilarial densities at intervals over this 9 month period. Given that the microfilaraemia is usually periodic (requiring blood samples to be taken in the night), it would not be trivial to evaluate the parasitological responses to ABZ if resistance is suspected. Furthermore, modelling of anthelmintic resistance development and spread in W. bancrofti (Schwab et al. Reference Schwab, Churcher, Schwab, Basáñez and Prichard2006, Reference Schwab, Churcher, Schwab, Basáñez and Prichard2007) indicates that frank parasitological evidence of anthelmintic resistance will not be obvious while the GPELF is active in a region, until treatment stops and microfilarial levels bounce back but are refractory to the existing drug combinations. For these reasons, there is a need for molecular markers for anthelmintic resistance in W. bancrofti and for monitoring to be conducted as part of the GPELF throughout the period of mass drug administration (MDA) as well as, for a period, after MDA has stopped.

Should we be concerned about anthelmintic resistance developing in nematode parasites of people?

It has been argued that anthelmintic resistance is unlikely to develop in lymphatic filariae for several reasons: (a) ABZ and MBZ have been used against soil transmitted nematodes of humans for many years and unequivocal proof of BZ-resistance is still lacking in any of them; (b) anthelmintic resistance has not arisen in animal filarial parasites, such as Dirofilaria immitis, which are biologically closely related to the human filarial nematodes; resistance problems in livestock have been in trichostrongylid nematodes, which are biologically different from the filariae; (c) filarial nematodes have a long life span and a generation interval of about one year, which would slow selection for resistance; (d) for LF, combination MDA with drugs with different modes of action are being used (ABZ+IVM, or ABZ+DEC) rather than monotherapy; and (e) in contrast to the frequent use of anthelmintics at short intervals in livestock, the MDA for LF is only once per year.

It is important to examine these points and others which might influence the rate of resistance development in the human filarial nematodes. With respect to point (a), there are some reports which suggest that resistance may have developed in a few instances in human soil transmitted nematodes (see above). Furthermore, investigators have not been looking for anthelmintic resistance in human nematode parasites, so we do not really know the extent to which it exists, if any. In addition, there has recently been a marked intensification of MDA for human helminth parasites as seen with the GPELF, the FRESH (Focusing Resources on Effective School Health, Hygiene & Nutrition) programme launched by WHO, UNICEF, UNESCO, World Bank and others in 2000 (see www.schoolsandhealth.org), APOC (African Program for Onchocerciasis Control) and many other large programmes. These large-scale MDA programmes in which a high degree of community coverage is sought could be putting strong pressure on the parasites to develop resistance. However, monitoring for drug resistance has not been a component of these programmes. In addition, it must be remembered that in the case of the human filarial parasites such as W. bancrofti, there are no free living stages which could provide refugia at the time of MDA. Almost all of the parasite population is in the people being treated, with only a very small portion of the total parasite population in the vector. Thus, almost the whole of the parasite population will be under drug selection pressure. This is in contrast to the soil transmitted nematodes of both humans and animals, where normally much of the total parasitic nematode population, at the time of treatment, will be on soil as the free-living egg or larval stages. Therefore, the selection pressures for resistance are likely to be much greater in the case of the human filarial nematodes than for soil transmitted nematodes of humans or animals.

Two important considerations are relevant to point (b). First, it is true that there are some biological differences between the trichostrongylid nematode parasites of livestock in which anthelmintic resistance has, in many cases, become a crisis, and the human filarial and gastrointestinal nematodes. However, we use the same anthelmintics and the available evidence strongly indicates that the receptors for them are similar in the human and livestock parasites. Furthermore, there are no a priori reasons why similar resistance mechanisms should not be expected in the different parasitic nematodes. Secondly, the belief that no resistance is occurring in the canine filarial nematode D. immitis is unsubstantiated by the lack of surveys designed to detect it. Despite the fact that IVM is used at a much lower dose rate as a prophylaxis against canine heartworm than in human filariases, the Food and Drug Administration in the USA is concerned about microfilarial breakthroughs in dogs that have been on ML prophylaxis (Hampshire, Reference Hampshire2005) as possible indicator of anthelmintic resistance development. While resistance remains to be confirmed in D. immitis, we cannot take comfort from the recent experience in this animal filarial infection. In addition to these considerations, there is evidence for non-random mating in filarial nematodes such as W. bancrofti and O. volvulus (Schwab et al. Reference Schwab, Churcher, Schwab, Basáñez and Prichard2007; Bourguinat, C., Pion, S. D. S., Kamgno, J., Gardon, J., Duke, B. O. L., Boussinesq, M. and Prichard, R. K., personal communication). BZ resistance in parasitic nematodes is usually recessive (Prichard, Reference Prichard2001) and if non-random mating typifies these infections, ABZ resistance will more rapidly develop than if mating was random, because non-random mating will increase the proportion of the population that is homozygous (see Schwab et al. Reference Schwab, Churcher, Schwab, Basáñez and Prichard2007 for further discussion).

With respect to point (c), it is true that filarial nematodes of people have a longer life span (estimates range from 5–8 years for W. bancrofti; 12–15 years for O. volvulus), in the absence of antiparasite chemotherapy, than most soil transmitted nematodes of animals or humans. Filariae typically require about 1 year between microfilarial production and achieving sexual maturity, compared with a minimal prepatent period for, e.g., trichostrongylid parasites of livestock (18 days). However, in reality, trichostrongylid nematodes of livestock will often only have 1 or 2 generations in a year. The longer life span and longer generation time for the human filariae will tend to slow the selection for anthelmintic resistance. However, it must also be recognized that existing anthelmintics used in MDA programmes are only slightly/moderately macrofilaricidal (see e.g. Duke et al. Reference Duke, Zea-Flores, Castro, Cupp and Munoz1990; Michael et al. Reference Michael, Malecela-Lazaro, Simonsen, Pedersen, Barker, Kumar and Kazura2004). The most important effect of MDA against the human filarial parasites is the prolonged suppression of reproduction by surviving adult worms. If some adult worms are more likely to survive repeated treatment and can resume reproduction more rapidly than other adult worms, these more tolerant (resistant?) worms will be producing progeny for a long time while other more susceptible adult worms will either be removed or have their reproduction almost totally suppressed throughout their life. This could result in a huge reproductive advantage for any resistant parasites and very strong selection pressure for resistance, and possibly a much stronger selective effect than in the case of parasites with a short prepatent period and a shorter life span than the human filariae.

In general, the use of combination chemotherapy, if the different drugs do not share the same mechanisms of resistance, should delay the development of drug resistance (point (d)). If drug resistance is entirely associated with changes in the drug receptor and the component drugs target different receptors, then their use in combination may be expected to delay the development of resistance to both. However, if the mechanism of resistance is in part due to non-receptor related factors, such as changes in the levels of ABC transporters, then the benefits of combination chemotherapy may not be realized, if similar changes can affect both drug classes. IVM is an excellent substrate for some P-glycoproteins and ABC transporters appear to be involved in IVM resistance (see Prichard and Roulet, in this special issue). BZ drugs such as ABZ can also be substrates for ABC transporters (Nare et al. Reference Nare, Liu, Prichard and Georges1994; Kerboeuf et al. Reference Kerboeuf, Blackhall, Kaminsky and von Samson-Himmelstjerna2003), so it is possible that combination MDA with ABZ+IVM could result in enhanced selection on ABC transporters involved with resistance to both classes of anthelmintics. In addition, it has recently been found that IVM selects on β-tubulin in both O. volvulus and H. contortus (Eng and Prichard, Reference Eng and Prichard2005; Eng et al. Reference Eng, Blackhall, Osei-Atweneboana, Bourguinat, Galazzo, Beech, Unnasch, Awadzi, Lubega and Prichard2006; Bourguinat, C., Pion, S. D. S., Kamgno, J., Gardon, J., Duke, B. O. L., Boussinesq, M. and Prichard, R. K., personal communication), which is known to be the receptor for ABZ. Therefore, it is possible that combination chemotherapy with ABZ+IVM may accelerate anthelmintic resistance selection rather than delay it. While this is speculative at this stage, it should be noted that the frequency of the Phe200Tyr SNP in β-tubulin in W. bancrofti microfilariae increased dramatically with one or two rounds of this combination treatment in patients in Burkina Faso (Schwab et al. Reference Schwab, Boakye, Kyelem and Prichard2005).

Finally, annual MDA for LF should impose less selection pressure for anthelmintic resistance than would the more frequent treatment regimens which are often used in livestock (point (e)). It should be noted that anthelmintic resistance has occurred rapidly in parasites of sheep even in cases in which only two treatments per year were used (Besier and Love, Reference Besier and Love2003). This occurred in a situation of a low level of refugia, the situation that prevails in LF when MDA is applied. Furthermore, as noted above, the anthelmintics used for LF MDA exert a prolonged effect over several months, whereas with livestock parasites the main effect is the rapid killing of all of the parasitic stages. The prolonged and continuous effect of ABZ, IVM (and DEC) against W. bancrofti could be expected to exert a very high level of selection pressure for resistance, possible exceeding that experienced when anthelmintics are used several times a year against soil transmitted nematodes.

For all of these compelling reasons, it would appear to be folly not to be concerned about anthelmintic resistance development in lymphatic filariae and not to monitor for resistance as part of the GPELF. However, as discussed above, it is not easy to monitor, parasitologically, for anthelmintic resistance in lymphatic filariae. For these reasons, DNA-based assays for anthelmintic resistance are urgently needed. Schwab et al. (Reference Schwab, Boakye, Kyelem and Prichard2005) have reported Phe200Tyr and Phe167Tyr SNP assays, using either Real Time PCR or Pyrosequencing, to monitor for the genetic changes most commonly associated with ABZ resistance in many nematodes. Monitoring for IVM resistance in lymphatic filariae could be developed based on understanding IVM resistance, and markers for resistance, in other nematodes (see McCavera et al. and Prichard and Roulet, in this special issue). The possibility of developing markers for DEC resistance is virtually non-existent at this time, as we do not even know how the drug works.

ABZ and MBZ are the main anthelmintics used to control soil transmitted nematodes in humans. As noted above, reports of treatment failures with BZs in these parasites remain to be confirmed. There is a critical need to develop assays to monitor for anthelmintic resistance in soil transmitted nematodes (Albonico et al. Reference Albonico, Engels and Savioli2004a). An in vitro test based on an assessment of drug effects on the mobility of human hookworm (Ancylostoma spp. and N. americanus) and Strongyloides spp. infective-stage larvae was developed to allow an estimate to be made of anthelmintic responsiveness (Kotze et al. Reference Kotze, Clifford, O'Grady, Behnke and McCarthy2004). However, such an assay would not readily work with parasites such as A. lumbricoides and T. trichiura, for which the infective stage is an egg. Recently, the hamster has been proposed as a laboratory animal model to examine anthelmintic resistance in human hookworms. Hamsters were infected with N. americanus and dosed with ABZ, MBZ and other anthelmintic drugs (Xue et al. Reference Xue, Hui-Qing, Jun-Ming, Fujiwara, Zhan, Hotez and Shu-Hua2005). These biological tests for anthelmintic resistance in some soil transmitted nematodes of people will be useful. Nevertheless, molecular markers for BZ resistance in soil transmitted nematodes are desirable and would enable monitoring to be more easily undertaken. In an attempt to determine whether the SNPs in β-tubulin which cause BZ resistance in nematodes also occur in T. trichiura and hookworms, this gene was sequenced from several isolates (Bennett et al. Reference Bennett, Anderson, Barker, Michael and Bundy2002; Albonico et al. Reference Albonico, Engels and Savioli2004b). The resistance associated SNPs was not detected in the relatively small number of worms sequenced. However, a more expansive investigation of these SNPs, coupled to field egg count monitoring and laboratory biological assays for resistance, is needed. Molecular assays similar to those developed for livestock parasites (see Samson-Himmelstjerna et al. in this special issue) or for W. bancrofti (Schwab et al. Reference Schwab, Boakye, Kyelem and Prichard2005) are needed for soil transmitted nematode parasites of humans.