Veterinary Parasitology, 35 (1990) 61-69 Elsevier Science Publishers B.V., Amsterdam - - Printed in The Netherlands
61
A N o v e l Approach to the Control of AnthelminticResistant H a e m o n c h u s c o n t o r t u s in Sheep J.A. VAN WYK 1 and P.C. VAN S C H A L K W Y K 2
*Veterinary Research Institute, Onderstepoort 0110 (South Africa) 2SmithKline Animal Health (Pty) Ltd., P.O. Box 38, Isando 1600 (South Africa) (Accepted for publication 28 June 1989)
ABSTRACT Van Wyk, J.A. and van Schalkwyk, P.C., 1990. A novel approach to the control of anthelminticresistant Haemonchus contortus in sheep. Vet. Parasitol., 35: 61-69. This paper reports attempts to control a resistant strain of Haemonchus contortus on pasture by replacing it with a susceptible strain. By making use of artificially infected donor sheep, six camps (paddocks) were seeded with a resistant field strain of H. contortus until it was confirmed by means of worm-free tracer lambs that the grazing had become infective. Thereafter, using donor sheep infected with a susceptible laboratory strain of H. contortus for seeding the pasture, attempts were made at various times of the year to replace the resistant strain on the pasture with the susceptible strain in five of the camps. The sixth remained as a control camp, in which no attempt was made to replace the resistant strain. In two of the five test camps, the susceptible strain was introduced in the autumn after 8-10 weeks of nil grazing; in the remaining three camps the introduction was made in spring (two camps) or summer without having a period of nil grazing. The susceptibility of the worm strains introduced initially, as well as of those that developed in the various camps, was gauged both by controlled non-parametric anthelmintic slaughter trials at the beginning and at the conclusion of the trial, and by an in vitro egg hatch test. A reversion to susceptibility occurred in three of the five camps. These included both of the camps infested with the susceptible strain in the spring and one of the two infested in the autumn. Possible reasons for the failure to replace the resistant strain in the remaining two test camps are discussed.
INTRODUCTION
Resistance of worms to the modern anthelmintics is reaching such proportions that some worm strains are already resistant to three of the four main anthelmintic groups in general use today (van Wyk et al., 1987; van Wyk and Malan, 1988). In Australia, milch goat farmers had to be informed that there were no longer any compounds available for controlling the resistant worm strains on their farms (Waller, 1985 ). At that stage, ivermectin and closantel
0304-4017/90/$03.50
© 1990 Elsevier Science Publishers B.V.
62
J.A. VAN WYK AND P.C. VAN SCALKWYK
had not been registered in Australia, but there is already resistance to both of these compounds in South Africa (van W y k and Malan, 1988). Thus it seems likely that they also will not prove to be a lasting solution to the problem of resistance that those farmers are experiencing. The realisation that new vermifuges are being developed at a rate that is too slow to keep abreast of the escalation of resistance has stimulated workers internationally to search for alternative methods to the exclusive use of anthelmintics in worm control, in order to conserve the efficacy of the compounds already in use and to protect them from being engulfed by the stream of resistance (Michel, 1976; Armour, 1983; Dineen, 1985; Hotson, 1985; Le Jambre, 1985). Unfortunately, relatively little progress has been made with the development of alternative methods of control, except for some dramatic results in the field of integrated worm control. In a few instances, worm control with integrated methods was so effective that there were problems in the moderate climate of Europe owing to excessive susceptibility of adult animals that were not exposed to sufficient worms while young (Armour, 1983). However the methods of integrated control used in Europe may not be equally effective in South Africa because of the long summers and differences in the local worm species. It is striking that few of the alternatives at present being investigated in the search for new methods of control appear to be aimed directly at the free-living stages of the worms on pasture. The common gastrointestinal nematodes of ruminants, being unable to multiply in their hosts, are dependent for survival on the ability of their free-living stages to develop and survive on pasture until they can gain access to susceptible hosts. Thus an excess of free-living stages is usually produced, and the vast majority of the total worm population usually occurs on the pasture and not in the host. The free-living stages on pasture can only migrate very short distances and nematodes are thus largely dependent upon the host for dissemination. The result is that the free-living stages are "cornered" on the pasture, unable to escape the effects of control measures if movement of the hosts is controlled. Furthermore, unless there is large-scale drainage of storm water, there should be little spill-over between properties if movement of animals is prevented. Thus, in contrast to what may be expected with other parasites, such as winged insects, few precautions should be necessary to ensure that a worm control programme will usually not be jeopardised by resistance on adjacent properties. This paper reports an experiment in which an attempt was made to overwhelm a resistant strain of Haemonchus contortus on pasture by seeding the grazing with a susceptible strain of the same worm species at different times of the year.
ANTHELMINTIC-RESISTANT HAEMONCHUS CONTORTUS CONTROL
63
MATERIALS AND METHODS
Experimental design The trial site was near K e m p t o n Park on the Transvaal Highveld. Over a period of 14 weeks, six camps (paddocks) consisting of unirrigated kikuyu pastures (Pennisetum clandestinum) were all seeded with the same benzimidazole-resistant strain of H. contortus. This was done by making use of donor sheep that were continually rotated at short intervals between all the camps, the time of seeding in each camp being more or less in relation to its size. The stocking rate during this period was ~ 7 sheep per hectare. W h e n it was confirmed by faecal egg counts of tracer lambs that the grazing had become infective, the various camps were randomly assigned to different treatments, as outlined in Table 1. Thereafter the stocking rate for each of the camps was maintained at ~ 16 sheep per hectare for the remainder of the trial. In order to maintain infection, 2-4 sheep infected with the resistant strain ran in each of the camps, according to size. Thereafter, with the exception of Camp 1, in which no a t t e m p t was made to substitute the resistant strain with one susceptible to the benzimidazole anthelmintics, these donor sheep were replaced at different times of the year with sheep infested with a susceptible strain of H. contortus. The latter sheep were maintained in the camps for the duration of the trial until mid-1987. In Camps 2 and 3, the donors of the susceptible strain were introduced 8 or 10 weeks after the donors of the resistant strain had been removed from the pasture. In the remaining three camps, the donor sheep were introduced on the same day that the donors of the resistant strain were removed; thus these camps were not rested. The strains of H. contortus that developed on the different pastures were isolated in the laboratory by means of worm-free tracer sheep that were grazed on the pastures for periods of 4 weeks. In the laboratory, the susceptibility of the various strains to the benzimidazole anthelmintics was tested (i) in controlled anthelmintic efficacy tests conducted at the start and at the conclusion of the trial and (ii) at various intervals by an in vitro egg hatch test (Hall et al., 1978). The latter test was modified as follows: a commercial formulation of thiabendazole [Thibenzole ( M S D ) ] was used in 4 log dilutions, varying from 1.47 to 0.00147 mg ml-1, for each test and the worm strains were compared by calculating the mean percentage hatch at the four concentrations of thiabendazole.
Animals and techniques The sheep used in the trial were of mixed breeds and sexes, and varied in age from 6 to 12 months. T h e y were dewormed on introduction to the laboratory,
64
J.A. VAN WYK AND P.C. VAN SCALKWYK
TABI,E 1 Trial design: exchange of donor sheep infected with a resistant strain of" H. contortus with others infected with a susceptible strain Camp
1 (Control) b 2 3 4 5 6
Period of nil grazing a (weeks)
Introduction of susceptible strain Season
Date
8 10 0c 0 0
Autumn Autumn Spring Spring Summer
28 11 10 29 17
M a r c h 1986 April 1986 September 1986 October 1986 December 1986
aNil grazing immediately before introduction of the susceptible strain of H. contortus. bControl camp, no a t t e m p t at replacing the resistant strain on pasture with the susceptible strain. CThe donors of the resistant strain were withdrawn a n d replaced on the same day by donors of the susceptible strain.
thereafter being housed in pens, on concrete, under conditions that precluded unintentional infection with worms. Immediately before being used in the trial, no worm ova could be demonstrated in 5 g of faeces using the flotation technique of Whitlock (1959). Maintenance of pure nematode strains in the laboratory, infection of trial animals, necropsy procedure, and worm recovery and identification were carried out as described by Reinecke (1973) with the following modifications: the ingesta were not concentrated by using a nylon cloth in a water bath and only a 1 × 1/10 aliquot of ingesta was examined per sample, except when total counts were required in those sheep that had worm counts critical for statistical analysis by the non-parametric method ( N P M ) of Groeneveld and Reinecke (1969). Allocation of animals and camps to treatment groups occurred randomly, with the aid of tables of random numbers. The numbers of sheep used in the N P M efficacy trials varied from 7 to 9 untreated controls and from 11 to 13 sheep in the drenched groups for each strain tested. The resistant strain, with which the pasture was initially seeded, and the susceptible strain were tested with both albendazole [Valbazen (SmithKline) ] at a dosage rate of 3.8 mg k g - 1 and thiabendazole [Thibenzole ( M S D ) ] at 44 mg kg -1 live mass. At the conclusion of the trial, the strains from each camp were tested only with albendazole at 3.8 mg k g - 1 live mass.
Statistical analysis Worm recovery data were analysed by means of the N P M of Groeneveld and Reinecke (1969), as modified by Clark (Reinecke, 1973). In addition, the per-
ANTHELMINTIC-RESISTANT HAEMONCHUS CONTORTUS CONTROL
65
centage efficacy was calculated based on comparison of the geometric mean worm burdens of the treated and untreated control sheep. In the latter calculations, a value of one was substituted for each nil value when no worms were recovered from an animal. RESULTS
While the worm burdens of the resistant strain of H. contortus that developed from larvae on the pastures from the initial seeding were reduced by only TABLE 2 Initial N P M trial: susceptibility of the strains of H. contortus to benzimidazoles Warm strain
Drench
Efficacy M e a n reduction ~ (%)
N P M class b
Field
Albendazole Thiabendazole
29.4 11.2
X X
Laboratory
Albendazole Thiabendazole
99.9 99.8
A A
aBased on a comparison of the geometric m e a n worm burdens of the treated and the untreated control sheep. bA signifies > 80% effective in > 80% of the treated flock; X signifies ineffective. TABLE 3 Susceptibility to albendazole of the strains of H. contortus in the various camps at the conclusion of the trial Camp No.
1 2 3 4 5 6
(Control) (Week + 8) ( W e e k + 10) (September 10) (October 29) (December 17)
M e a n worm burden a
Efficacy
Untreated
Treated
M e a n reduction (%)a
N P M class b
1147 1828 925 1373 956 1937
833 71 437 55 4 1019
27.4 96.1 52.7 96.0 99.6 47.4
X A X A A X
aGeometric mean. bBased on a comparison of the geometric mean worm burdens of the treated and the untreated control sheep. CEfficacy classification by the n o n - p a r a m e t r i c method of Groeneveld and Reinecke (1969); A denotes > 80% effective in > 80% of the treated flock; X signifies ineffective.
61
A N o v e l Approach to the Control of AnthelminticResistant H a e m o n c h u s c o n t o r t u s in Sheep J.A. VAN WYK 1 and P.C. VAN S C H A L K W Y K 2
*Veterinary Research Institute, Onderstepoort 0110 (South Africa) 2SmithKline Animal Health (Pty) Ltd., P.O. Box 38, Isando 1600 (South Africa) (Accepted for publication 28 June 1989)
ABSTRACT Van Wyk, J.A. and van Schalkwyk, P.C., 1990. A novel approach to the control of anthelminticresistant Haemonchus contortus in sheep. Vet. Parasitol., 35: 61-69. This paper reports attempts to control a resistant strain of Haemonchus contortus on pasture by replacing it with a susceptible strain. By making use of artificially infected donor sheep, six camps (paddocks) were seeded with a resistant field strain of H. contortus until it was confirmed by means of worm-free tracer lambs that the grazing had become infective. Thereafter, using donor sheep infected with a susceptible laboratory strain of H. contortus for seeding the pasture, attempts were made at various times of the year to replace the resistant strain on the pasture with the susceptible strain in five of the camps. The sixth remained as a control camp, in which no attempt was made to replace the resistant strain. In two of the five test camps, the susceptible strain was introduced in the autumn after 8-10 weeks of nil grazing; in the remaining three camps the introduction was made in spring (two camps) or summer without having a period of nil grazing. The susceptibility of the worm strains introduced initially, as well as of those that developed in the various camps, was gauged both by controlled non-parametric anthelmintic slaughter trials at the beginning and at the conclusion of the trial, and by an in vitro egg hatch test. A reversion to susceptibility occurred in three of the five camps. These included both of the camps infested with the susceptible strain in the spring and one of the two infested in the autumn. Possible reasons for the failure to replace the resistant strain in the remaining two test camps are discussed.
INTRODUCTION
Resistance of worms to the modern anthelmintics is reaching such proportions that some worm strains are already resistant to three of the four main anthelmintic groups in general use today (van Wyk et al., 1987; van Wyk and Malan, 1988). In Australia, milch goat farmers had to be informed that there were no longer any compounds available for controlling the resistant worm strains on their farms (Waller, 1985 ). At that stage, ivermectin and closantel
0304-4017/90/$03.50
© 1990 Elsevier Science Publishers B.V.
62
J.A. VAN WYK AND P.C. VAN SCALKWYK
had not been registered in Australia, but there is already resistance to both of these compounds in South Africa (van W y k and Malan, 1988). Thus it seems likely that they also will not prove to be a lasting solution to the problem of resistance that those farmers are experiencing. The realisation that new vermifuges are being developed at a rate that is too slow to keep abreast of the escalation of resistance has stimulated workers internationally to search for alternative methods to the exclusive use of anthelmintics in worm control, in order to conserve the efficacy of the compounds already in use and to protect them from being engulfed by the stream of resistance (Michel, 1976; Armour, 1983; Dineen, 1985; Hotson, 1985; Le Jambre, 1985). Unfortunately, relatively little progress has been made with the development of alternative methods of control, except for some dramatic results in the field of integrated worm control. In a few instances, worm control with integrated methods was so effective that there were problems in the moderate climate of Europe owing to excessive susceptibility of adult animals that were not exposed to sufficient worms while young (Armour, 1983). However the methods of integrated control used in Europe may not be equally effective in South Africa because of the long summers and differences in the local worm species. It is striking that few of the alternatives at present being investigated in the search for new methods of control appear to be aimed directly at the free-living stages of the worms on pasture. The common gastrointestinal nematodes of ruminants, being unable to multiply in their hosts, are dependent for survival on the ability of their free-living stages to develop and survive on pasture until they can gain access to susceptible hosts. Thus an excess of free-living stages is usually produced, and the vast majority of the total worm population usually occurs on the pasture and not in the host. The free-living stages on pasture can only migrate very short distances and nematodes are thus largely dependent upon the host for dissemination. The result is that the free-living stages are "cornered" on the pasture, unable to escape the effects of control measures if movement of the hosts is controlled. Furthermore, unless there is large-scale drainage of storm water, there should be little spill-over between properties if movement of animals is prevented. Thus, in contrast to what may be expected with other parasites, such as winged insects, few precautions should be necessary to ensure that a worm control programme will usually not be jeopardised by resistance on adjacent properties. This paper reports an experiment in which an attempt was made to overwhelm a resistant strain of Haemonchus contortus on pasture by seeding the grazing with a susceptible strain of the same worm species at different times of the year.
ANTHELMINTIC-RESISTANT HAEMONCHUS CONTORTUS CONTROL
63
MATERIALS AND METHODS
Experimental design The trial site was near K e m p t o n Park on the Transvaal Highveld. Over a period of 14 weeks, six camps (paddocks) consisting of unirrigated kikuyu pastures (Pennisetum clandestinum) were all seeded with the same benzimidazole-resistant strain of H. contortus. This was done by making use of donor sheep that were continually rotated at short intervals between all the camps, the time of seeding in each camp being more or less in relation to its size. The stocking rate during this period was ~ 7 sheep per hectare. W h e n it was confirmed by faecal egg counts of tracer lambs that the grazing had become infective, the various camps were randomly assigned to different treatments, as outlined in Table 1. Thereafter the stocking rate for each of the camps was maintained at ~ 16 sheep per hectare for the remainder of the trial. In order to maintain infection, 2-4 sheep infected with the resistant strain ran in each of the camps, according to size. Thereafter, with the exception of Camp 1, in which no a t t e m p t was made to substitute the resistant strain with one susceptible to the benzimidazole anthelmintics, these donor sheep were replaced at different times of the year with sheep infested with a susceptible strain of H. contortus. The latter sheep were maintained in the camps for the duration of the trial until mid-1987. In Camps 2 and 3, the donors of the susceptible strain were introduced 8 or 10 weeks after the donors of the resistant strain had been removed from the pasture. In the remaining three camps, the donor sheep were introduced on the same day that the donors of the resistant strain were removed; thus these camps were not rested. The strains of H. contortus that developed on the different pastures were isolated in the laboratory by means of worm-free tracer sheep that were grazed on the pastures for periods of 4 weeks. In the laboratory, the susceptibility of the various strains to the benzimidazole anthelmintics was tested (i) in controlled anthelmintic efficacy tests conducted at the start and at the conclusion of the trial and (ii) at various intervals by an in vitro egg hatch test (Hall et al., 1978). The latter test was modified as follows: a commercial formulation of thiabendazole [Thibenzole ( M S D ) ] was used in 4 log dilutions, varying from 1.47 to 0.00147 mg ml-1, for each test and the worm strains were compared by calculating the mean percentage hatch at the four concentrations of thiabendazole.
Animals and techniques The sheep used in the trial were of mixed breeds and sexes, and varied in age from 6 to 12 months. T h e y were dewormed on introduction to the laboratory,
64
J.A. VAN WYK AND P.C. VAN SCALKWYK
TABI,E 1 Trial design: exchange of donor sheep infected with a resistant strain of" H. contortus with others infected with a susceptible strain Camp
1 (Control) b 2 3 4 5 6
Period of nil grazing a (weeks)
Introduction of susceptible strain Season
Date
8 10 0c 0 0
Autumn Autumn Spring Spring Summer
28 11 10 29 17
M a r c h 1986 April 1986 September 1986 October 1986 December 1986
aNil grazing immediately before introduction of the susceptible strain of H. contortus. bControl camp, no a t t e m p t at replacing the resistant strain on pasture with the susceptible strain. CThe donors of the resistant strain were withdrawn a n d replaced on the same day by donors of the susceptible strain.
thereafter being housed in pens, on concrete, under conditions that precluded unintentional infection with worms. Immediately before being used in the trial, no worm ova could be demonstrated in 5 g of faeces using the flotation technique of Whitlock (1959). Maintenance of pure nematode strains in the laboratory, infection of trial animals, necropsy procedure, and worm recovery and identification were carried out as described by Reinecke (1973) with the following modifications: the ingesta were not concentrated by using a nylon cloth in a water bath and only a 1 × 1/10 aliquot of ingesta was examined per sample, except when total counts were required in those sheep that had worm counts critical for statistical analysis by the non-parametric method ( N P M ) of Groeneveld and Reinecke (1969). Allocation of animals and camps to treatment groups occurred randomly, with the aid of tables of random numbers. The numbers of sheep used in the N P M efficacy trials varied from 7 to 9 untreated controls and from 11 to 13 sheep in the drenched groups for each strain tested. The resistant strain, with which the pasture was initially seeded, and the susceptible strain were tested with both albendazole [Valbazen (SmithKline) ] at a dosage rate of 3.8 mg k g - 1 and thiabendazole [Thibenzole ( M S D ) ] at 44 mg kg -1 live mass. At the conclusion of the trial, the strains from each camp were tested only with albendazole at 3.8 mg k g - 1 live mass.
Statistical analysis Worm recovery data were analysed by means of the N P M of Groeneveld and Reinecke (1969), as modified by Clark (Reinecke, 1973). In addition, the per-
ANTHELMINTIC-RESISTANT HAEMONCHUS CONTORTUS CONTROL
65
centage efficacy was calculated based on comparison of the geometric mean worm burdens of the treated and untreated control sheep. In the latter calculations, a value of one was substituted for each nil value when no worms were recovered from an animal. RESULTS
While the worm burdens of the resistant strain of H. contortus that developed from larvae on the pastures from the initial seeding were reduced by only TABLE 2 Initial N P M trial: susceptibility of the strains of H. contortus to benzimidazoles Warm strain
Drench
Efficacy M e a n reduction ~ (%)
N P M class b
Field
Albendazole Thiabendazole
29.4 11.2
X X
Laboratory
Albendazole Thiabendazole
99.9 99.8
A A
aBased on a comparison of the geometric m e a n worm burdens of the treated and the untreated control sheep. bA signifies > 80% effective in > 80% of the treated flock; X signifies ineffective. TABLE 3 Susceptibility to albendazole of the strains of H. contortus in the various camps at the conclusion of the trial Camp No.
1 2 3 4 5 6
(Control) (Week + 8) ( W e e k + 10) (September 10) (October 29) (December 17)
M e a n worm burden a
Efficacy
Untreated
Treated
M e a n reduction (%)a
N P M class b
1147 1828 925 1373 956 1937
833 71 437 55 4 1019
27.4 96.1 52.7 96.0 99.6 47.4
X A X A A X
aGeometric mean. bBased on a comparison of the geometric mean worm burdens of the treated and the untreated control sheep. CEfficacy classification by the n o n - p a r a m e t r i c method of Groeneveld and Reinecke (1969); A denotes > 80% effective in > 80% of the treated flock; X signifies ineffective.
