- Email: [email protected]
Reproductive potential of Echinococcus multilocularis in experimentally infected foxes, dogs, raccoon dogs and cats
International Journal for Parasitology 36 (2006) 79–86 www.elsevier.com/locate/ijpara
Reproductive potential of Echinococcus multilocularis in experimentally infected foxes, dogs, raccoon dogs and cats C.M.O. Kapel a, P.R. Torgerson b, R.C.A. Thompson c, P. Deplazes a,b,* a
WHO/FAO Collaborating Centre for Parasitic Zoonoses, Danish Centre for Experimental Parasitology, Royal Veterinary and Agricultural University, Dyrlaegevej 100, DK 1870 Frederiksberg C, Denmark b WHO Collaborating Centre for Parasitic Zoonoses, Institute of Parasitology, University of Zurich, Winterthurerstrasse 266A, 8057 Zurich, Switzerland c WHO Collaborating Centre for the Molecular Epidemiology of Parasitic Infections, School of Veterinary and Biomedical Sciences, Murdoch University, 6150 Murdoch, WA, Australia Received 4 May 2005; received in revised form 16 August 2005; accepted 23 August 2005
Abstract A total of 15 red foxes, 15 raccoon dogs, 15 domestic dogs and 15 domestic cats were each infected with 20,000 protoscolices of Echinococcus multilocularis. At 35, 63, and 90 days post inoculation (dpi), five animals from each group were necropsied and the worm burdens determined. The highest worm burdens in foxes (mean of 16,792) and raccoon dogs (mean of 7930) were found at 35 dpi. These declined to a mean of just 331 worms in foxes and 3213 worms in raccoon dogs by day 63 with a further decline to 134 worms in foxes and 67 worms in raccoon dogs by day 90. In dogs, there was no significant difference between worm burdens recovered at days 35 (mean of 2466) and day 90 (mean of 1563), although reduced numbers were recovered on day 63 (mean of 899). In cats, worms were found in four animals 35 dpi (mean of 642), in three at 63 dpi (mean of 28) and in two at 90 dpi (mean of 57). Faecal egg counts were determined at 3 day intervals from 25 dpi. A mathematical model of egg excretion dynamics suggested that the mean biotic potential per infected animal was high in foxes (346,473 eggs); raccoon dogs (335,361 eggs) and dogs (279,910 eggs) but very low for cats (573 eggs). It also indicated that approximately 114, 42 and 27 eggs per worm were excreted in the faeces of dogs, raccoon dogs and foxes, respectively. The fecundity of worms in cats was low with an average of less than one egg per worm. The peak levels of coproantigen were detected earlier in foxes and raccoon dogs than in dogs. Eggs recovered from foxes, raccoon dogs and dogs resulted in massive infections in experimental mice. However, metacestodes did not develop from eggs originating from infected cats. It is concluded that foxes, raccoon dogs and dogs are good hosts of E. multilocularis. In contrast, the low worm establishment, the very few excreted eggs and the lack of infectivity of eggs strongly indicate that cats play an insignificant role in parasite transmission. q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved. Keywords: Echinococcus multilocularis; Experimental infection; Dogs; Racoon dogs; Foxes; Cats; Biotic potential; Patent period
1. Introduction The life-cycle of Echinococcus multilocular predominantly involves foxes (Vulpes vulpes in temperate regions and Alopex lagopus in arctic and sub-arctic regions) as definitive hosts and many species of rodents can act as intermediate hosts (Schantz et al., 1995). Foxes are thought to be the main sources of environmental contamination with eggs of E. multilocularis in most endemic areas (Eckert and Deplazes, 2004), but infections have also been found in coyotes (Canis latrans), * Corresponding author. Address: Institute of Parasitology, University of Zurich, Winterthurerstr 266A, CH-8057 Zurich, Switzerland. Tel.: C411 635 85 02; fax: C411 635 89 07. E-mail address: [email protected] (P. Deplazes).
raccoon-dogs (Nyctereutes procyonoides), wolves (Canis lupus) and occasionally wild cats (Felis libyca) (Vuitton et al., 2003). However, it is not known if any other species can maintain the E. multilocularis cycle independently from the fox reservoir. In certain areas, a relatively high prevalence of E. multilocularis has been found in domestic dogs (1–12%). These include endemic foci in Alaska, China and Central Europe (Schantz et al., 1995; Deplazes et al., 2004; Budke et al., 2005). In the past, many such studies could only be based on necropsy results, with the disadvantage that the animals examined represented a selected population. More recent studies from China were based on the use of arecoline purgation (Budke et al., 2005) but have the disadvantage of the undefined diagnostic sensitivity of this technique.
0020-7519/$30.00 q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.ijpara.2005.08.012
80
C.M.O. Kapel et al. / International Journal for Parasitology 36 (2006) 79–86
Using a coproantigen-ELISA in combination with a confirmatory PCR for ELISA positive results, larger numbers of dogs have recently been investigated in Europe. Deplazes et al. (1999) recorded a prevalence of 0.3% (95% confidence interval (CI) 0.04–1.09%) whilst Gottstein et al. (2001) found up to 6.9% (95% CI 2.60–14.57%) in Swiss dog populations. Although dogs may not maintain a synanthropic cycle of E. multilocularis in Europe, even a low prevalence may constitute a zoonotic risk due to their proximity to humans (Kern et al., 2004). Prevalence of E. multilocularis in cat populations, as determined at necropsy, ranged between 0.5 and 3.7% in various endemic areas (Eckert and Deplazes, 2004) but generally only low worm burdens and variable development of the proglottids have been detected. Although small numbers of E. multilocularis eggs from faeces of naturally infected cats have been characterized by PCR (Deplazes et al., 1999; Gottstein et al., 2001), the infectivity of such eggs have never been proven. The raccoon dog (N. procyonoides) originated in Eastern Asia and was introduced as a game and farm animal into the Baltic part of the former Soviet Union in the period 1930–1955 and has gradually spread to Western Europe along the Baltic Sea coast (Pruˆsaite` et al., 1988; Asikainen et al., 2003). First reports of E. multilocularis infection in raccoon dogs were from Japan and recently Thiess et al. (2001) found E. multilocularis infection in raccoon dogs in Germany. In a recent study in Japan, three of 15 (23%) raccoon dogs were found to be infected with an average of 505 worms (Yimam et al., 2002). Thus, only very limited information about parasite numbers and development is available and the epidemiological significance of the raccoon dog as a definitive host of E. multilocularis is unknown. Because of the establishment of this carnivore in new areas of Europe, such information is of public health significance. In this study, the dynamics of E. multilocularis worm burden, excretion of eggs and coproantigens were compared in experimentally infected foxes, raccoon dogs, dogs and cats. 2. Materials and methods 2.1. Experimental animals Fifteen red foxes (V. vulpes, six male, nine female) and 15 raccoon dogs (N. procyonoides, 10 male, five female) obtained from a large scale Danish fur farm (Møldrup minkfarm), in addition to 15 dogs (FBI hounds, six male, nine female) and 15 cats (European short hair, seven male, eight female) bred for experimental purposes by Harland (Dogs: Harland Sprague Dawley, Madison, Wisconsin, USA; Cats: Harland Nederland, Horst, Netherlands) were used for the study. Vixens of the foxes and raccoon dogs were treated with fenbendazole 5 weeks before birth and the cubs were treated 8 weeks after birth (50 mg/kg body weight on three consecutive days). Dogs and cats were certified helminth free by the breeding company. The animals were kept in the experimental facility for 2 weeks prior to the study to allow them to adapt to the new surroundings,
routine and diet. All animals were fed a heat-treated meal (approx: 75% fish meal, 25% rye flour) once daily and examined weekly by a veterinarian. All animals were 13–15 weeks of age at the time of inoculation where the average weight was 1.6 kg for foxes, 2.9 kg for raccoon dogs, 14.6 kg for dogs, and 2.1 kg for cats. The animals were kept under Danish experimental animal permission no: 2000-561-321 and treated in accordance with animal ethics laws of the EU. 2.2. Echinococcus multilocularis isolate The E. multilocularis isolate (EM280) used in this study was originally obtained from a naturally infected common vole (Arvicola terrestris) from Zurich and subsequently passaged in the laboratory (i.p. inoculation of 0.1 ml of minced metacestode material) in jirds (Meriones unguiculatus) and field voles (Microtus arvalis) for less than 6 months. Metacestodes from the laboratory rodents were isolated and stored for around 18 h in phosphate-buffered saline (pH 7.2) containing penicillin (100 U/ml) and streptomycin (100 mg/ml) at 4 8C. The metacestode tissues were pooled and minced by scissors and protoscolices isolated through a sieve (mesh size of 1 mm) 3 h before inoculation. The viability of the protoscolices was examined by the activity of the flame cells (around 90% of the protoscolices examined were active). Prior to inoculation, the cats, dogs, foxes and raccoon dogs were anaesthetized by IM injection (0.1 ml/5 kg body weight) of a mixture of 1 ml saline, 250 mg ZOLETIL 50w (125 mg zolazepamC125 mg tiletamin, Boehringer Ingelheim), and 2.5 ml DOMITOR VETw (medetomidin (1 mg/ml), Orion Pharma). Each experimental host received 20,000 protoscolices by stomach intubation. Two h after inoculation, a full portion of their normal fish-meal feed was administered to the animals. 2.3. Sample collection and processing Faecal samples were collected daily from the bottom of the cages and stored at K80 8C for at least 3 days before further processing. Quantitative faecal egg counts (epg), were subsequently undertaken at 3 day intervals from 25 days post inoculation (dpi) on a 4 g homogenised faecal sample by McMaster glucose flotation (Roepstorff and Nansen, 1998). The excreted faecal mass was recorded during the first week p.i. to allow for calculation of the total egg excretion. Specific E. multilocularis coproantigen detection was performed as described by Deplazes et al. (1999). The cut-off value for the estimation of specific reactions for each animal species was determined with the mean plus three standard deviations of the ELISA values of the samples collected on the day of infection. From each animal species, five animals, chosen at random, were sacrificed on each of 35, 63 and 90 dpi by heart puncture (0.5 ml/kg pentobarbital (20%)) under anaesthesia (see above). Within the first 30 min after necropsy of the individual animal, the intestine was incised longitudinally and cut into three equal segments. Each intestinal segment was incubated for 30 min at 37 8C in 0.5 L of physiological saline solution. Worm recovery
C.M.O. Kapel et al. / International Journal for Parasitology 36 (2006) 79–86
and processing were conducted according to Hofer et al. (2000). 2.4. Infectivity of eggs To evaluate their infectivity, eggs were recovered from fresh faeces 35 dpi (foxes, raccoon dogs and dogs) as described by Mathis et al. (1996). Due to their very low numbers in the faeces of cats, eggs were additionally collected from gravid proglottids 63 dpi (foxes, raccoon dogs, and cats). The infectivity was tested by oral inoculation in Naval Medical Research Institute (NMRI) mice (6 weeks of age). A dose of 2000 eggs was inoculated by a feeding needle in each of five mice (2800 eggs from worms of cats in each of two mice) (Table 2). After 6 months, the entire body and internal organs were carefully examined under dissection microscope at 40! magnification for the presence of metacestodes. 2.5. Biosecurity precautions All parts of the experiment were conducted at containment level 3 (EU council directive 90/679/EEC). Thus, the housing and laboratories were isolated from any other activities, marked with ‘biohazard’ and only accessible for staff dedicated to the activities. The wearing of particle barrier masks and protective clothing was undertaken. Decontamination and showering procedures were mandatory on leaving contaminated areas. All surfaces were cleaned by sodium hypochlorite (2%). Infective material was placed in double containers: after being placed in the first container which was surface treated with sodium hypochlorite (2%) and then placed in a second clean container. All waste and animals carcasses were incinerated after the experiment. The housing was high pressure cleaned twice with a 5% solution of hypochlorite and left for 2 weeks before any other activity was permitted. 2.6. Analysis It was not possible to sample each animal every day. Therefore, to estimate the total biotic potential per infected host and per worm it is necessary to model the faecal egg excretion dynamics. A starting point in this process would be the cumulative faecal egg output at time t. This will be sigmoidal with two asymptotes. The lower asymptote is 0 as initially (in the prepatent period) the egg output is 0. The upper asymptote will represent the parasite’s biotic potential as this will represent the total cumulative egg production over the lifetime of the infection. This can be modelled by a sigmoidal growth curve determined by a series of logistic equations given by NðtÞ Z N0 C
NN 1 C expða½bKtÞ
(1)
where N(t)is the quantity of interest at time t, in this case the cumulative faecal egg count, N0 is the faecal egg count before patency (i.e. Z0), NN is the upper asymptote (i.e. the biotic potential), a is the curvature parameter and b is a constant.
81
As the initial egg count is 0, N0Z0, which reduces Eq. (1) to three parameters. To increase the flexibility of this to model variations in the slope of the curve, an extra parameter can be introduced to give: NðtÞ Z
NN f1 C expða½bKtÞgc
(2)
Eq. (2) is the Richard’s growth curve (Richards, 1959). As egg counts were taken from experimental animals at various time points Eq. (2) needs to be manipulated to analyse the daily egg output. The rate of change of the total cumulative egg output, dN(t)/dt is the instantaneous rate of egg production, and can be approximated as the daily faecal egg output M(t). This can be derived by differentiating (2) and shown to be: MðtÞ Z
NNac expða½bKtÞ f1 C expða½bKtÞgcC1
(3)
Because egg counts were not taken every day, due to the difficult and dangerous nature of the experimental conditions, the cumulative egg count N(t) cannot be calculated directly. However, the daily egg production for each animal can be calculated from the epg and the average daily mass of faeces that was produced. Although some animals were removed from the study 35 and 63 dpi, all the data generated from these animals was used in modelling the dynamics of egg output. Because the variance of the mean of the daily faecal egg output was much greater than the mean, the variation in daily egg output on each day was then fitted to the model described by Eq. (2) using maximum likelihood techniques assuming the epg followed a negative binomial distribution (for a more detailed description of the technique see Torgerson et al., 2003). A likelihood profile of the variation of the parameter values, given the data and the model was found by using the likelihood profile function of the Excele (Microsoft Corp, Redmond WA) PopTools add on (CSIRO Australia). From this likelihood profile confidence intervals of the parameter values were calculated (Venzon and Moolgavkar, 1988). Confidence intervals of the faecal egg output were constructed using Monte-Carlo techniques (Torgerson et al., 2003). Goodness of fit was estimated using the parametric bootstrap technique (Efron and Tibshirani, 1993). Likewise bootstrap techniques were used to determine if any of the observed data points contribute to lack of fit in the model. Significant differences between parameters in different groups of animals were determined by comparing the likelihood of the model with a single set of parameters for the combined data set with those with separate parameters (Hilbourn and Mangel, 1997). Worm burdens between different groups of animals sacrificed 35, 63 and 90 dpi were also compared using maximum likelihood techniques and a negative binomial model of parasite aggregation. The likelihood of several groups with different means was compared to groups with a single mean (Hilbourn and Mangel, 1997) and probabilities calculated based on the likelihood ratio test. Point estimates of the time to excrete 50, 75 and 95% of the mean total egg excretion were made by substituting the most likely values of NN, a, b and c and the relevant values of N(t)
82
C.M.O. Kapel et al. / International Journal for Parasitology 36 (2006) 79–86
and solving Eq. (1) for t. To estimate confidence intervals for t, the defined distributions of N8, a, b and c were sampled 10,000 times across the distributions calculated from their likelihood profile. This generated 10,000 values for t, and the 250th and 9750th ranked samples were taken as the lower and upper 95% confidence limits, respectively. Statistical differences between times were based on the rank percentile that the mean time to excrete the relevant percentage burden occurred in the comparison group. Estimates of the mean egg excretion per parasite were made by calculating the overall egg production (NN) and dividing this by the worm burden isolated at 35 dpi for each host species. Confidence intervals were estimated by Monte-Carlo techniques using the defined distribution of worm burden and biotic potential calculated from the likelihood profile. The probability of differences between egg productions of worms obtained from different host species was calculated in a similar manner. 3. Results Worms established in the intestines of all host species, but infection frequency and worm burden varied significantly (Table 1). Worms were found in foxes and raccoon dogs at all Table 1 Worm burdens of Echinococcus multilocularis found at necropsy in cats, dogs, raccoon dogs, and foxes (nZ5) experimentally inoculated with 20,000 protoscolices
Individual worm burdens at 35 dpi
Mean worm burden Lower 95% CI Upper 95% CI Individual worm burdens at 63 dpi
Mean worm burden Lower 95% CI Upper 95% CI Individual worm burdens at 90 dpi
Mean worm burden Lower 95% CI Upper 95% CI
Fox
Raccoon dogs
Dog
Cat
16,580 16,244 11,732 19,704 19,700 16,792a 13,877 20,000 720 100 380 399 55 331f 157 889 63 10 419 160 17 134f 49 579
13,700 5,910 2,020 5,040 12,980 7930b 4365 16,703 3310 2390 7230 1180 1956 3,213d,f 1,820 6,476 140 110 20 28 35 67f,g 33 159
3238 3330 1925 2905 930 2466c 1609 4053 85 85 193 3 4130 899f 232 10,208 0 60 1,785 5,420 550 1,563d 332 20,000
1250 25 1690 0 244 642 144 11,748 120 0 5 16 0 28e,f 5 1,429 285 2 0 0 0 57 4 Indeterminate
dpi, days post inoculation; CI, confidence interval. a Significantly higher than in racoon dogs, dogs and cats at this time point. b Significantly higher than dogs and cats at this time point. c Significantly higher than cats at this time point. d Significantly higher than in the three other species at this time point. e Significantly lower than the three other species at this time point. f Significantly less than at 35 dpi in the same species. g Significantly less than at 63 dpi in the same species.
times, and in 14 of 15 dogs (one was negative at 90 dpi). In cats, worms were detected in four of five at 35 dpi, three animals at 63 dpi and two animals at 90 dpi. No clinical signs, which could be assigned to the parasite infection, were observed in any animals during the experiment. Individual worm burdens are summarised in Table 1. At 35 dpi, mean worm burdens representing 84, 39, 12, and 3.2% of protoscolices inoculated were counted in foxes, raccoon dogs, dogs and cats, respectively. Significant differences between worm burdens of different species at different time points are indicated in Table 1. There were no significant differences between egg counts from female and male animals and so data was pooled for further analysis. A number of alternative models were fitted to the egg count data. These included the sigmoidal models: the Verhulst growth curve and the models described by Eqs. (1)– (3). The best fit was the Richard’s growth curve with four parameters (Eqs. (2) and (3)) which gave a significant fit to the data from dogs, racoon dogs and foxes (P!0.005, P!0.01, P!0.025). Simple polynomial or exponential growth model models tended to gave a poorer fit (data not shown). The egg excretion dynamics, based on the results of the Richard’s model for all four species is reported in Table 2. There were no significant differences between the total biotic potential between foxes, racoon dogs and dogs. Figs. 1–3 display the model fit, the 95% confidence intervals together with the observed data for foxes, dogs and raccoon dogs The cats had much lower total egg output and egg excretion was sporadic. In total, only four eggs were detected in a single faecal sample 28 dpi whilst a single egg was detected in three other samples from different animals at 34–54 dpi. Hence, model parameters could not be derived with any accuracy for this species. The egg excretion, determined as epg x mean faecal mass, varied significantly over time and between the host species (Figs. 1–3). The average daily faecal masses were determined as 14.9 g for the foxes, 13.1 g for the raccoon dogs, 68.9 g for the dogs and 10.7 g for the cats. The model suggested that the maximum likelihood estimates (MLE) for the time to excretion of 50, 75 and 95% of the biotic potential was shortest in the fox and longest in the dog. However, the confidence intervals show that these differences were not significant (Table 3). The small numbers of eggs produced in infected cats were eliminated early (Table 3). The 95% CIs of the effective patent period, defined as the time taken for 95% of the eggs to be excreted Table 2 Maximum likelihood estimates and their 95% confidence intervals (CIs) of the biotic potential (total eggs excreted) following experimental inoculation with 20,000 protoscolices of Echinococcus multilocularis
Fox Raccoon dog Dog Cat
Biotic potential
CIs
346,473 335,361 279,910 573
230,420–581,130 206,490–578,819 183,559–463,719 Indeterminatea
The biotic potential in the cat was significantly lower than the other three species. a Gave a poor fit to the model.
Number of eggs per animal
Number of eggs per animal
C.M.O. Kapel et al. / International Journal for Parasitology 36 (2006) 79–86
A
40000 30000 20000 10000 0
0
10
20
30
40
50
60
70
80
A
40000 30000 20000 10000 0
0
0.80
A405nm
A405nm
0.40 0.20 0
10
20
30
40
50
60
70
80
40
50
60
70
80
10
20 30 40 50 60 70 Days after experimental infection
90
B
0.40
0.00
90
Fig. 1. (A) Dynamics of Echinococcus multilocularis egg excretion in faeces following experimental inoculation of 20,000 protoscolices in foxes. Observed data (C) (mean of recorded data at respective time points). Solid line is best fit to logistic model (–) and 95% confidence intervals (/) (P!0.01). Observed data indicated as (B) had evidence of lack of fit to the model. (B) Observed dynamics of detected coproantigen levels measured by ELISA in the faeces (GSD) of foxes.
was 17–44 days for foxes, 22–47 days for raccoon dogs and 22–93 days for dogs (Table 3). An estimate of 13 days was made for cats, but CIs could not be calculated for this species. The model suggested that the highest egg output occurred in Number of eggs per animal
30
0.20 0
Days after experimental infection
A
40000 30000 20000 10000 0 0
10
20
30
40
50
60
70
80
90
B
0.80 0.60 A 405 nm
20
0.60
0.60
0.40 0.20 0.00
10
90
B
0.80
0.00
83
0
10
20 30 40 50 60 70 Days after experimental infection
80
90
Fig. 2. (A) Dynamics of Echinococcus multilocularis egg excretion in faeces following experimental inoculation of 20,000 protoscolices in raccoon dogs. Observed data (C) (mean of recorded data at respective time points). Solid line is best fit to logistic model (–) and 95% confidence intervals (/) (P!0.025). Observed data indicated as (B) had evidence of lack of fit to the model. (B) Observed dynamics of detected coproantigen levels measured by ELISA in the faeces (GSD) of racoon dogs.
80
90
Fig. 3. (A) Dynamics of Echinococcus multilocularis egg excretion in faeces following experimental inoculation of 20,000 protoscolices in dogs. Observed data (C) (mean of recorded data at respective time points). Solid line is best fit to logistic model (–) and 95% confidence intervals (/) (p!0.005). (B) Observed dynamics of detected coproantigen levels measured by ELISA in the faeces (GSD) of dogs.
foxes 37–42 dpi, in raccoon dogs 38–43 dpi and in dogs 43– 45 dpi (Figs. 1–3). The highest daily egg excretion determined were 98,638 for a fox (40 dpi), 88,032 for a raccoon dog (45 dpi), 112,996 eggs for a dog (60 dpi) and 856 for a cat (28 dpi). The dynamics of coproantigen excretion for foxes, raccoon dogs and dogs are shown in Figs. 1–3. In all four animal species, specific reaction was found only sporadically within the first 10 dpi. Although specific coproantigen reactions were observed for cats from 21 dpi until the end of the experiment, around 50% of the samples gave high non-specific reactions, which made the results uninterpretable (data not shown). During 10–15 dpi, 75% of samples from foxes, 94% of samples from raccoon dogs and 31% samples from dogs were coproantigen positive. During 20–28 dpi, 100% of samples from both foxes and racoon dogs were positive and 86% of dog samples were positive. During the first part of patency (29–63 dpi), 82, 79 and 91% of samples from foxes, racoon dogs and dogs were coproantigen positive. In second part of patency, the figures were 51, 17 and 52%, respectively. Table 3 Time taken in days (G95% confidence intervals (CIs)) to expel 50, 75 and 95% of the total eggs excreted in cats, dogs, raccoon dogs and foxes experimentally inoculated with 20,000 protoscolices of Echinococcus multilocularis Proportion of total egg mass
50%
75%
95%
Fox Raccoon dog Dog Cat
9.9 (2.7–22.7) 15.9 (9.4–28.8) 18.8 (3.2–55.6) 6.7a
13.4 (5.7–27.1) 18.8 (12.0–32.5) 23.8 (7.1–63.4) 10.0a
26.8 (17.0–44.2) 30.3 (22.1–46.8) 43.0 (21.9–93.1) 13.0a
a
Indeterminate CIs.
84
C.M.O. Kapel et al. / International Journal for Parasitology 36 (2006) 79–86
Table 4 Infectivity in mice of Echinococcus multilocularis eggs isolated from either faeces or from gravid worms of experimentally infected dogs, raccoon dogs and foxes (number of isolated eggs from faeces of cats was too low for testing) Hosts
Source of eggs
Age of infection (dpi)
Dose per mouse
No. mice developing cysts
Foxes Raccoon dogs Dogs Foxes Raccoon dogs Cats
Faeces Faeces Faeces Worms Worms Worms
35 35 35 63 63 63
2000 2000 2000 2000 2000 2800
5 5 5 5 5 0
In each case, five experimental mice were infected, except with eggs from worms in cats, where only two mice were infected. dpi, days post inoculation.
The bioassay evaluating the infectivity of excreted eggs (Table 4) revealed that all mice inoculated with eggs isolated 35 dpi from the faeces of foxes, raccoon dogs and dogs developed metacestodes with fully developed protoscolices 6 months later. The low epg in cats did not allow for inoculation of mice with eggs from faeces. Eggs, which as an alternative were isolated directly from gravid proglottids of raccoon dogs and foxes at 63 dpi, resulted in fully developed metacestodes whereas eggs from proglottids of cats did not establish. The estimated mean number of eggs excreted per parasite for foxes was 27 (95% CIs 17–44), for raccoon dogs it was 42 (19–85), for dogs it was 114 (63–215), whilst for cats it was only 0.76 (0.02–22). Worms from dogs produced significantly more eggs than those from foxes (P!0.001) and raccoon dogs (P!0.01). There was no significant difference in the mean number of eggs produced per worm between foxes and raccoon dogs. 4. Discussion The experimental infections described provide important information on the biotic potential of E. multilocularis, the establishment of the infection, the dynamics of egg excretion, and the life-span of the infection. It is important to note that the data presented may not be applicable to other definitive hosts such as the arctic fox (A. lagopus) which may have different dynamics of infection. Foxes have a higher initial worm establishment with a shorter parasite life expectancy compared with the racoon dog and dog. This results in an earlier peak of egg production with a shorter period of high egg production, (Table 1, Figs. 1–3). In naturally infected foxes, worm burdens were found to be both highly over-dispersed and to be significantly higher in juvenile foxes compared with adults (Hofer et al., 2000; Raoul et al., 2001; Yimam et al., 2002). Because the dynamics of the infection in foxes appears to be characterised by a rapid loss of the worm burden, despite high initial numbers, foxes with low worm numbers are more likely to be sampled in a given study. However other factors such as different predation habits of the different age groups of foxes, host immunity or genetic susceptibility could also contribute to this phenomenon. Protoscolex numbers in naturally infected voles have also
been shown to be over-dispersed (Stieger et al., 2002). All the experimental animals were young. Hence, it is possible that the biotic potential calculated overestimates the true biotic potential in naturally exposed and possibly less susceptible adult animals. However, it is the highly infected individuals carrying thousands of fertile worms that are responsible for most of the egg contamination. Furthermore, Budke et al. (2005) reported there were no age-related changes in abundance in naturally infected dogs. There are few other data concerning the egg production and dynamics of worm burdens during infection with E. multilocularis. Nonaka et al. (1996) reported that in foxes inoculated with about 150,000 protoscolices, eggs were regularly detected in faeces from 29–60 dpi and sporadically until 84 dpi. Nevertheless, 11 and 91 parasites were detected in two animals 125 dpi. Lukashenko (1975) recovered 70,000 worms (from an infecting dose of 100,000 protoscolices) from dogs 3–4 weeks after inoculation, but found only 528 worms after 3–4 months and just 27 after 6 months. Interestingly, in the study of Vogel (1957) worm burdens ranging from less than 10 specimens to several thousand were found in dogs 42–97 dpi (with 0–55% gravid worms). However, these experiments were performed with different isolates of E. multilocularis and no information on protoscolex numbers was given. In the present study, significant numbers of worms were found at 90 dpi (Table 1) although very few eggs were found in the faeces at this time (Fig. 3). Cats appear to be poor definitive hosts with just 3.2% of the infecting dose establishing (Table 1). Likewise, Lukashenko (1975) recovered 7% of an infecting dose of 50,000 protoscolices 3–4 weeks after infection and just 0.3% after 3 months. Jenkins and Romig (2000, 2003) and Thompson et al. (2003) also recovered only 6.9, 3 and 12.7%, respectively, of the infecting dose of between 10,000 and 22,600 protoscolices between 21 and 25 dpi. Reports of naturally infected cats are characterised by low infection intensity. Petavy et al. (2000) found three of 81 necropsied cats naturally infected but with a mean intensity of just nine worms per cat. Deplazes et al. (1999) described one of 130 stray cats infected with just five parasites. A variety of mathematical equations were assessed to model the egg production dynamics and Eq. (3) gave the most satisfactory fit for the egg count data for foxes, racoon dogs and dogs. The poor fit for cats was probably due to the lack of data resulting from very low egg production. The model fit is consistent with egg production starting slowly at the start of patency, accelerating as the worms mature and then slowing as the worms become senescent. In the fox and raccoon dog, there was some evidence of lack of fit due to two and three data points, respectively (Figs. 1 and 2). Visual inspection suggests that these model outliers could be due to a second, but lower, peak in egg production. For E. multilocularis it has never been convincingly shown that sequential proglottids even develop to gravidity. The relatively short life span of the worms in foxes in this experiment supports the hypothesis that sequential proglottidisation did not contribute significantly to total egg production in this study. However, the model outliers in the fox and racoon
C.M.O. Kapel et al. / International Journal for Parasitology 36 (2006) 79–86
dog (Figs. 1 and 2) could indicate a biphasic response and hence a second wave of proglottid shedding. In dogs, the much higher egg production per worm, and the longer survival of the worms suggest that more than one proglottid per worm may contribute to egg counts but there was little evidence of a biphasic response in the model. Further studies should focus on the developmental processes of growth, segmentation, proglottidisation and maturation in adult Echinococcus to develop a better understanding of its reproductive biology (Thompson et al., unpublished data). The high biotic potential of infections from raccoon dogs demonstrates that the parasite in this host is of particular concern as the range of this species is expanding in Europe. Likewise, the high biotic potential in dogs demonstrates that these hosts could make a significant contribution to the transmission cycle. In some regions such as China, Alaska or Central Europe, dogs may have high prevalence rates (ranging from 7 to O20%) (Schantz et al., 1995; Gottstein et al., 2001; Budke et al., 2005), but even in the general dog population of Switzerland where there is a low prevalence of just 0.3%, the lifetime incidence of infection in dogs could be 10% (Deplazes et al., 2004). In addition, the fact that dogs can carry contamination with eggs on their fur and their close proximity to humans potentiate the zoonotic importance of this host species. Epidemiological studies in China and in Europe have suggested dog ownership and/or close contact with dogs are important risk factors for human alveolar echinococcosis (Craig et al., 2000; Wang et al., 2001; Kern et al., 2004). Descriptions of patent infections in naturally infected cats have been described in the single infected cat reported by Deplazes et al. (1999) and in one cat of 33 examined by Gottstein et al. (2001). No studies of naturally infected cats have demonstrated parasite abundance levels or levels of egg excretion that could be considered important in epidemiological terms. Nevertheless, because the experimental design of the present study necessarily involved relatively few animals, the results cannot absolutely rule out the cat as a possible host for parasite transmission. Although a similar biotic potential was recorded in three host species (Tables 1 and 2) differences in worm establishment rate resulted in a higher biotic potential per parasite in dogs (114) compared with foxes (27) and raccoon dogs (42). The high worm fecundity in dogs could result from a crowding effect caused by the high burdens in foxes compared with dogs. In a limited experiment, not analysed statistically, Heath and Lawrence (1991) suggested that three dogs experimentally inoculated with 88,000 protoscolices of E. granulosus reached patency earlier with a more rapid increase in faecal egg production compared with inoculations of 44,000 protoscolices. Alternatively, the immunological response to E. multilocularis may be more efficient in foxes resulting in more rapid expulsion of worms. The estimated numbers of eggs excreted per worm in faeces is low in comparison with the 200 eggs per gravid proglottid reported by Vogel (1957). This difference may be because a large number of non-viable eggs disintegrate in the intestine of the definitive hosts: the egg viability, measured in vitro by the resistance of onchospheres to sodium
85
hypochlorite, was only about 46% in eggs isolated from gravid proglottids (Deplazes et al., 2005). This compares with more than 99% for eggs isolated from faeces of the same hosts. It is also likely that faecal egg counts were underestimated. In a limited comparative study with the sieving technique (Mathis et al., 1996), the method used in this study had an egg recovery rate of approximately 70% (data not shown). In foxes, the peak levels of coproantigen production were a few days earlier than the peak egg production (Fig. 2) with levels declining from 35–45 dpi. This dynamic is similar to that recorded by Nonaka et al. (1996) and appears to mirror the decline in egg production and numbers of worms recovered at necropsy. Likewise, Nonaka et al. (1996) recovered very small numbers of adult worms at the end of their experimental infections. The peak coproantigen production in foxes could correspond to the most metabolically active stage of the infection with high worm burdens just before the onset of patency and during the main elimination of proglottids and worms. Thus, the coproantigen test will have a higher likelihood of detecting infection in the late prepatent and early patent phases of infection. This is important, as the early patent period in foxes appears to be the most significant in terms of transmission. Likewise, the coproantigen production in raccoon dogs showed high levels in the late prepatent period, but also a second peak at about 60 dpi (Fig. 2). Whilst the first peak can be interpreted in a similar manner to that in the foxes, the second peak may be an expression of loss of worms following loss of the terminal egg containing proglottids. In the dog (Fig. 3), the early peak in coproantigen production was absent and maximum levels were seen around 50–60 dpi, corresponding to the observed peak in epg and the extended persistence of the intestinal worms. For foxes, raccoon dogs and dogs, coproantigen levels were detected in a high proportion of samples during the second half of pre-patency and the first 30 days of patency confirming the usefulness of this tool in identifying infected animals. Although coproantigen was detected in cats (data not shown), this was at low and intermittent levels and hence no conclusions can be drawn in this species. Acknowledgements The Danish Veterinary and Agricultural Research Council and the Danish National Research Foundation are thanked for substantial financial support to the present study. The competent staff at the research facility “Lindholm” of the Danish Institute for Food and Veterinary Research is thanked for hard work, flexibility and logistical support. References Asikainen, J., Mustonen, A., Hyva¨rinen, H., Nieminen, P., 2003. Seasonal reproduction endocrine profile of the raccoon dog (Nyctereutes procyonoides)—effects of melatonin and food deprivation. J. Exp. Zoo. Part A: Comp. Exp. Biol. 299, 180–187. Budke, C.M., Craig, P.S., Jiamin, Q., Torgerson, P.R., 2005. Modeling the transmission of Echinococcus multilocularis and Echinococcus granulosus in dogs for a high endemic region of the Tibetan plateau. Int. J. Parasitol. 35, 163–170.
86
C.M.O. Kapel et al. / International Journal for Parasitology 36 (2006) 79–86
Craig, P.S., Giraudoux, P., Shi, D., Bartholomat, B., Barnish, G., Delattre, P., Quere, J.P., Harraga, S., Bao, G., Wang, Y., Lu, F., Ito, A., Vuitton, D.A., 2000. An epidemiological and ecological study of human alveolar echinococcosis transmission in south Gansu. China. Act. Trop. 77, 167–177. Deplazes, P., Alther, P., Tanner, I., Thompson, R.C.A., Eckert, J., 1999. Echinococcus multilocularis coproantigen detection by enzyme-linked immunosorbent assay in fox, dog, and cat populations. J. Parasitol. 85, 115–121. Deplazes, P., Hegglin, D., Gloor, S., Romig, T., 2004. Wilderness in the city: the urbanisation of Echinococcus multilocularis. Trends Parasitol. 20, 77–84. Deplazes, P., Grimm, F., Sydler, T., Tanner, I., Kapel, C.M.O., 2005. Experimental alveolar echinococcosis in pigs, lesion development and serological follow up. Vet. Parasitol. 130, 213–222. Eckert, J., Deplazes, P., 2004. Biological, epidemiological and clinical aspects of echinococcosis, a zoonosis of increasing concern. Clin. Microbiol. Rev. 17, 107–135. Efron, B., Tibshirani, R.J., 1993. An Introduction to the Bootstrap. Chapman & Hall, New York, NY. Gottstein, B., Saucy, F., Deplazes, P., Reichen, J., Demierre, G., Busato, A., Zuercher, C., Pugin, P., 2001. Is high prevalence of Echinococcus multilocularis in wild and domestic animals associated with disease incidence in humans? Emerg. Infect. Dis. 7, 408–412. Heath, D.D., Lawrence, S.B., 1991. Daily egg production of dogs infected with Echinococcus granulosus. Arch. Int. Hidatid. 30, 321–328. Hilbourn, R., Mangel, M., 1997. The Ecological Detective: Confronting Models with Data Monographs in Population Biology, 28. Princeton University Press, New Jersey. Hofer, S., Gloor, S., Mu¨ller, U., Mathis, A., Hegglin, D., Deplazes, P., 2000. High prevalence of Echinococcus multilocularis in urban red foxes (Vulpes vulpes) and voles (Arvicola terrestris) in the city of Zurich, Switzerland. Parasitology 120, 135–142. Jenkins, D.J., Romig, T., 2000. Eficacy of DroncitR spot-on (praziquantel) 4% w/v against immature and mature Echinococcus multilocularis in cats. Int. J. Parasitol. 30, 959–962. Jenkins, D.J., Romig, T., 2003. Milbemycin oxime in a new formulation combined with praziquantel, does not reduce the efficacy of praziquantel against Echinococcus multilocularis in cats. J. Helminthol. 77, 367–370. Kern, P., Ammon, A., Kron, M., Sinn, G., Sander, S., Petersen, L.R., Gaus, W., Kern, P., 2004. Risk factors for alveolar echinococcosis in human. Emerg. Infect. Dis. 10, 2088–2093. Lukashenko, N.P., 1975. Alveokokkoz, Meditsina, Moskva 1975. 327 pp. (In Russian). Mathis, A., Deplazes, P., Eckert, J., 1996. Improved test system for PCR-based specific detection of Echinococcus multilocularis eggs. J. Helminthol. 70, 219–222. Nonaka, N., Iida, M., Yagi, K., Ito, T., Ooi, H.K., Oku, Y., Kamiya, M., 1996. Time course of coproantigen excretion in Echinococcus multilocularis
infections in foxes and an alternative definitive host, golden hamsters. Int. J. Parasitol. 26, 1271–1278. Petavy, A.F., Tenora, F., Deblock, S., Sergent, V., 2000. Echinococcus multilocularis in domestic cats in France. A potential risk factor for alveolar hydatid disease contamination in humans. Vet. Parasitol. 87, 151–156. Pruˆsaite`, J., Mazeikyte, R., Pauza, D., 1988. Fauna of Lithuania (Lietuvos Fauna). Mokslav, Vilnius, p. 186–218. Raoul, F., Deplazes, P., Nonaka, N., Piarroux, R., Vuitton, D.A., Giraudoux, P., 2001. Assessment of the epidemiological status of Echinococcus multilocularis in foxes in France using ELISA coprotests on fox faeces collected in the field. Int. J. Parasitol. 31, 1579–1588. Richards, F.J., 1959. A flexible growth function for empirical use. J. Exp. Bot. 10, 290–300. Roepstorff, A., Nansen, P., 1998. The Epidemiology. Diagnosis and Control of Helminth Parasites of Swine FAO Animal Health Manual No. 3. FAO, Rome, Italy. Schantz, P.M., Chai, J., Craig, P.A., Eckert, J., Jenkins, D.J., Macpherson, C.N.L., Thakur, A., 1995. Epidemiology and control of hydatid disease. In: Thompson, R.C.A., Lymbery, A.J. (Eds.), Echinococcus and hydatid disease. CAB International, Wallingford, UK, pp. 233–331. Stieger, C., Hegglin, D., Schwarzenbach, G., Mathis, A., Deplazes, P., 2002. Spatial and temporal aspects of urban transmission of Echinococcus multilocularis. Parasitology 124, 631–640. Thiess, A., Schuster, R., Nockler, K., Mix, H., 2001. Helminth findings in indigenous raccoon dogs Nyctereutes procyonoides (Gray, 1843).. Berl. Mu¨nch. Tierarztl. Wschr. [German] 114, 176–273. Thompson, R.C.A., Deplazes, P., Eckert, J., 2003. Observations on the development of Echinococcus multilocularis in cats. J. Parasitol. 89, 1086–1088. Torgerson, P.R., Shaikenov, B.S., Rysmukhambetova, A.T., Abdybekova, A.M., Usenbayev, A.E., Baitursinov, K.K., 2003. Modelling the transmission dynamics of Echinococcus granulosus in dogs in rural Kazakhstan. Parasitology 126, 417–424. Venzon, D.J., Moolgavkar, S.H., 1988. A method for computing profilelikelihood-based confidence intervals. Appl. Stat. 37, 87–94. Vogel, H., 1957. Echinococcus multilocularis in South Germany. I. The tapeworm stage of strains from humans and animals. Z. Tropenmed. Parasitol. 8, 404–454. Vuitton, D.A., Zhou, H., Bresson-Hadni, S., Wang, Q., Piarroux, M., Raoul, F., Giraudoux, P., 2003. Epidemiology of alveola echinococcosis with particular reference to China and Europe. Parasitology 127, S87–S107. Wang, Q., Qiu, J., Schantz, P., He, J.G., Ito, A., Liu, F., 2001. Investigation of risk factors for development of human hydatidosis among households raising livestock in Tibetan areas of western Sichuan Province. Chin. J. Parasitol. Parasit. Dis. 19, 93–96. Yimam, A.E., Nonaka, N., Oku, Y., Kamiya, M., 2002. Prevalence and intensity of Echinococcus multilocularis in red foxes (Vulpes vulpes schrencki) and raccoon dogs (Nyctereutes procyonoides albus) in Otaru City, Hokkaido, Japan. Jpn. J. Vet. Res. 49, 287–296.
Reproductive potential of Echinococcus multilocularis in experimentally infected foxes, dogs, raccoon dogs and cats C.M.O. Kapel a, P.R. Torgerson b, R.C.A. Thompson c, P. Deplazes a,b,* a
WHO/FAO Collaborating Centre for Parasitic Zoonoses, Danish Centre for Experimental Parasitology, Royal Veterinary and Agricultural University, Dyrlaegevej 100, DK 1870 Frederiksberg C, Denmark b WHO Collaborating Centre for Parasitic Zoonoses, Institute of Parasitology, University of Zurich, Winterthurerstrasse 266A, 8057 Zurich, Switzerland c WHO Collaborating Centre for the Molecular Epidemiology of Parasitic Infections, School of Veterinary and Biomedical Sciences, Murdoch University, 6150 Murdoch, WA, Australia Received 4 May 2005; received in revised form 16 August 2005; accepted 23 August 2005
Abstract A total of 15 red foxes, 15 raccoon dogs, 15 domestic dogs and 15 domestic cats were each infected with 20,000 protoscolices of Echinococcus multilocularis. At 35, 63, and 90 days post inoculation (dpi), five animals from each group were necropsied and the worm burdens determined. The highest worm burdens in foxes (mean of 16,792) and raccoon dogs (mean of 7930) were found at 35 dpi. These declined to a mean of just 331 worms in foxes and 3213 worms in raccoon dogs by day 63 with a further decline to 134 worms in foxes and 67 worms in raccoon dogs by day 90. In dogs, there was no significant difference between worm burdens recovered at days 35 (mean of 2466) and day 90 (mean of 1563), although reduced numbers were recovered on day 63 (mean of 899). In cats, worms were found in four animals 35 dpi (mean of 642), in three at 63 dpi (mean of 28) and in two at 90 dpi (mean of 57). Faecal egg counts were determined at 3 day intervals from 25 dpi. A mathematical model of egg excretion dynamics suggested that the mean biotic potential per infected animal was high in foxes (346,473 eggs); raccoon dogs (335,361 eggs) and dogs (279,910 eggs) but very low for cats (573 eggs). It also indicated that approximately 114, 42 and 27 eggs per worm were excreted in the faeces of dogs, raccoon dogs and foxes, respectively. The fecundity of worms in cats was low with an average of less than one egg per worm. The peak levels of coproantigen were detected earlier in foxes and raccoon dogs than in dogs. Eggs recovered from foxes, raccoon dogs and dogs resulted in massive infections in experimental mice. However, metacestodes did not develop from eggs originating from infected cats. It is concluded that foxes, raccoon dogs and dogs are good hosts of E. multilocularis. In contrast, the low worm establishment, the very few excreted eggs and the lack of infectivity of eggs strongly indicate that cats play an insignificant role in parasite transmission. q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved. Keywords: Echinococcus multilocularis; Experimental infection; Dogs; Racoon dogs; Foxes; Cats; Biotic potential; Patent period
1. Introduction The life-cycle of Echinococcus multilocular predominantly involves foxes (Vulpes vulpes in temperate regions and Alopex lagopus in arctic and sub-arctic regions) as definitive hosts and many species of rodents can act as intermediate hosts (Schantz et al., 1995). Foxes are thought to be the main sources of environmental contamination with eggs of E. multilocularis in most endemic areas (Eckert and Deplazes, 2004), but infections have also been found in coyotes (Canis latrans), * Corresponding author. Address: Institute of Parasitology, University of Zurich, Winterthurerstr 266A, CH-8057 Zurich, Switzerland. Tel.: C411 635 85 02; fax: C411 635 89 07. E-mail address: [email protected] (P. Deplazes).
raccoon-dogs (Nyctereutes procyonoides), wolves (Canis lupus) and occasionally wild cats (Felis libyca) (Vuitton et al., 2003). However, it is not known if any other species can maintain the E. multilocularis cycle independently from the fox reservoir. In certain areas, a relatively high prevalence of E. multilocularis has been found in domestic dogs (1–12%). These include endemic foci in Alaska, China and Central Europe (Schantz et al., 1995; Deplazes et al., 2004; Budke et al., 2005). In the past, many such studies could only be based on necropsy results, with the disadvantage that the animals examined represented a selected population. More recent studies from China were based on the use of arecoline purgation (Budke et al., 2005) but have the disadvantage of the undefined diagnostic sensitivity of this technique.
0020-7519/$30.00 q 2005 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.ijpara.2005.08.012
80
C.M.O. Kapel et al. / International Journal for Parasitology 36 (2006) 79–86
Using a coproantigen-ELISA in combination with a confirmatory PCR for ELISA positive results, larger numbers of dogs have recently been investigated in Europe. Deplazes et al. (1999) recorded a prevalence of 0.3% (95% confidence interval (CI) 0.04–1.09%) whilst Gottstein et al. (2001) found up to 6.9% (95% CI 2.60–14.57%) in Swiss dog populations. Although dogs may not maintain a synanthropic cycle of E. multilocularis in Europe, even a low prevalence may constitute a zoonotic risk due to their proximity to humans (Kern et al., 2004). Prevalence of E. multilocularis in cat populations, as determined at necropsy, ranged between 0.5 and 3.7% in various endemic areas (Eckert and Deplazes, 2004) but generally only low worm burdens and variable development of the proglottids have been detected. Although small numbers of E. multilocularis eggs from faeces of naturally infected cats have been characterized by PCR (Deplazes et al., 1999; Gottstein et al., 2001), the infectivity of such eggs have never been proven. The raccoon dog (N. procyonoides) originated in Eastern Asia and was introduced as a game and farm animal into the Baltic part of the former Soviet Union in the period 1930–1955 and has gradually spread to Western Europe along the Baltic Sea coast (Pruˆsaite` et al., 1988; Asikainen et al., 2003). First reports of E. multilocularis infection in raccoon dogs were from Japan and recently Thiess et al. (2001) found E. multilocularis infection in raccoon dogs in Germany. In a recent study in Japan, three of 15 (23%) raccoon dogs were found to be infected with an average of 505 worms (Yimam et al., 2002). Thus, only very limited information about parasite numbers and development is available and the epidemiological significance of the raccoon dog as a definitive host of E. multilocularis is unknown. Because of the establishment of this carnivore in new areas of Europe, such information is of public health significance. In this study, the dynamics of E. multilocularis worm burden, excretion of eggs and coproantigens were compared in experimentally infected foxes, raccoon dogs, dogs and cats. 2. Materials and methods 2.1. Experimental animals Fifteen red foxes (V. vulpes, six male, nine female) and 15 raccoon dogs (N. procyonoides, 10 male, five female) obtained from a large scale Danish fur farm (Møldrup minkfarm), in addition to 15 dogs (FBI hounds, six male, nine female) and 15 cats (European short hair, seven male, eight female) bred for experimental purposes by Harland (Dogs: Harland Sprague Dawley, Madison, Wisconsin, USA; Cats: Harland Nederland, Horst, Netherlands) were used for the study. Vixens of the foxes and raccoon dogs were treated with fenbendazole 5 weeks before birth and the cubs were treated 8 weeks after birth (50 mg/kg body weight on three consecutive days). Dogs and cats were certified helminth free by the breeding company. The animals were kept in the experimental facility for 2 weeks prior to the study to allow them to adapt to the new surroundings,
routine and diet. All animals were fed a heat-treated meal (approx: 75% fish meal, 25% rye flour) once daily and examined weekly by a veterinarian. All animals were 13–15 weeks of age at the time of inoculation where the average weight was 1.6 kg for foxes, 2.9 kg for raccoon dogs, 14.6 kg for dogs, and 2.1 kg for cats. The animals were kept under Danish experimental animal permission no: 2000-561-321 and treated in accordance with animal ethics laws of the EU. 2.2. Echinococcus multilocularis isolate The E. multilocularis isolate (EM280) used in this study was originally obtained from a naturally infected common vole (Arvicola terrestris) from Zurich and subsequently passaged in the laboratory (i.p. inoculation of 0.1 ml of minced metacestode material) in jirds (Meriones unguiculatus) and field voles (Microtus arvalis) for less than 6 months. Metacestodes from the laboratory rodents were isolated and stored for around 18 h in phosphate-buffered saline (pH 7.2) containing penicillin (100 U/ml) and streptomycin (100 mg/ml) at 4 8C. The metacestode tissues were pooled and minced by scissors and protoscolices isolated through a sieve (mesh size of 1 mm) 3 h before inoculation. The viability of the protoscolices was examined by the activity of the flame cells (around 90% of the protoscolices examined were active). Prior to inoculation, the cats, dogs, foxes and raccoon dogs were anaesthetized by IM injection (0.1 ml/5 kg body weight) of a mixture of 1 ml saline, 250 mg ZOLETIL 50w (125 mg zolazepamC125 mg tiletamin, Boehringer Ingelheim), and 2.5 ml DOMITOR VETw (medetomidin (1 mg/ml), Orion Pharma). Each experimental host received 20,000 protoscolices by stomach intubation. Two h after inoculation, a full portion of their normal fish-meal feed was administered to the animals. 2.3. Sample collection and processing Faecal samples were collected daily from the bottom of the cages and stored at K80 8C for at least 3 days before further processing. Quantitative faecal egg counts (epg), were subsequently undertaken at 3 day intervals from 25 days post inoculation (dpi) on a 4 g homogenised faecal sample by McMaster glucose flotation (Roepstorff and Nansen, 1998). The excreted faecal mass was recorded during the first week p.i. to allow for calculation of the total egg excretion. Specific E. multilocularis coproantigen detection was performed as described by Deplazes et al. (1999). The cut-off value for the estimation of specific reactions for each animal species was determined with the mean plus three standard deviations of the ELISA values of the samples collected on the day of infection. From each animal species, five animals, chosen at random, were sacrificed on each of 35, 63 and 90 dpi by heart puncture (0.5 ml/kg pentobarbital (20%)) under anaesthesia (see above). Within the first 30 min after necropsy of the individual animal, the intestine was incised longitudinally and cut into three equal segments. Each intestinal segment was incubated for 30 min at 37 8C in 0.5 L of physiological saline solution. Worm recovery
C.M.O. Kapel et al. / International Journal for Parasitology 36 (2006) 79–86
and processing were conducted according to Hofer et al. (2000). 2.4. Infectivity of eggs To evaluate their infectivity, eggs were recovered from fresh faeces 35 dpi (foxes, raccoon dogs and dogs) as described by Mathis et al. (1996). Due to their very low numbers in the faeces of cats, eggs were additionally collected from gravid proglottids 63 dpi (foxes, raccoon dogs, and cats). The infectivity was tested by oral inoculation in Naval Medical Research Institute (NMRI) mice (6 weeks of age). A dose of 2000 eggs was inoculated by a feeding needle in each of five mice (2800 eggs from worms of cats in each of two mice) (Table 2). After 6 months, the entire body and internal organs were carefully examined under dissection microscope at 40! magnification for the presence of metacestodes. 2.5. Biosecurity precautions All parts of the experiment were conducted at containment level 3 (EU council directive 90/679/EEC). Thus, the housing and laboratories were isolated from any other activities, marked with ‘biohazard’ and only accessible for staff dedicated to the activities. The wearing of particle barrier masks and protective clothing was undertaken. Decontamination and showering procedures were mandatory on leaving contaminated areas. All surfaces were cleaned by sodium hypochlorite (2%). Infective material was placed in double containers: after being placed in the first container which was surface treated with sodium hypochlorite (2%) and then placed in a second clean container. All waste and animals carcasses were incinerated after the experiment. The housing was high pressure cleaned twice with a 5% solution of hypochlorite and left for 2 weeks before any other activity was permitted. 2.6. Analysis It was not possible to sample each animal every day. Therefore, to estimate the total biotic potential per infected host and per worm it is necessary to model the faecal egg excretion dynamics. A starting point in this process would be the cumulative faecal egg output at time t. This will be sigmoidal with two asymptotes. The lower asymptote is 0 as initially (in the prepatent period) the egg output is 0. The upper asymptote will represent the parasite’s biotic potential as this will represent the total cumulative egg production over the lifetime of the infection. This can be modelled by a sigmoidal growth curve determined by a series of logistic equations given by NðtÞ Z N0 C
NN 1 C expða½bKtÞ
(1)
where N(t)is the quantity of interest at time t, in this case the cumulative faecal egg count, N0 is the faecal egg count before patency (i.e. Z0), NN is the upper asymptote (i.e. the biotic potential), a is the curvature parameter and b is a constant.
81
As the initial egg count is 0, N0Z0, which reduces Eq. (1) to three parameters. To increase the flexibility of this to model variations in the slope of the curve, an extra parameter can be introduced to give: NðtÞ Z
NN f1 C expða½bKtÞgc
(2)
Eq. (2) is the Richard’s growth curve (Richards, 1959). As egg counts were taken from experimental animals at various time points Eq. (2) needs to be manipulated to analyse the daily egg output. The rate of change of the total cumulative egg output, dN(t)/dt is the instantaneous rate of egg production, and can be approximated as the daily faecal egg output M(t). This can be derived by differentiating (2) and shown to be: MðtÞ Z
NNac expða½bKtÞ f1 C expða½bKtÞgcC1
(3)
Because egg counts were not taken every day, due to the difficult and dangerous nature of the experimental conditions, the cumulative egg count N(t) cannot be calculated directly. However, the daily egg production for each animal can be calculated from the epg and the average daily mass of faeces that was produced. Although some animals were removed from the study 35 and 63 dpi, all the data generated from these animals was used in modelling the dynamics of egg output. Because the variance of the mean of the daily faecal egg output was much greater than the mean, the variation in daily egg output on each day was then fitted to the model described by Eq. (2) using maximum likelihood techniques assuming the epg followed a negative binomial distribution (for a more detailed description of the technique see Torgerson et al., 2003). A likelihood profile of the variation of the parameter values, given the data and the model was found by using the likelihood profile function of the Excele (Microsoft Corp, Redmond WA) PopTools add on (CSIRO Australia). From this likelihood profile confidence intervals of the parameter values were calculated (Venzon and Moolgavkar, 1988). Confidence intervals of the faecal egg output were constructed using Monte-Carlo techniques (Torgerson et al., 2003). Goodness of fit was estimated using the parametric bootstrap technique (Efron and Tibshirani, 1993). Likewise bootstrap techniques were used to determine if any of the observed data points contribute to lack of fit in the model. Significant differences between parameters in different groups of animals were determined by comparing the likelihood of the model with a single set of parameters for the combined data set with those with separate parameters (Hilbourn and Mangel, 1997). Worm burdens between different groups of animals sacrificed 35, 63 and 90 dpi were also compared using maximum likelihood techniques and a negative binomial model of parasite aggregation. The likelihood of several groups with different means was compared to groups with a single mean (Hilbourn and Mangel, 1997) and probabilities calculated based on the likelihood ratio test. Point estimates of the time to excrete 50, 75 and 95% of the mean total egg excretion were made by substituting the most likely values of NN, a, b and c and the relevant values of N(t)
82
C.M.O. Kapel et al. / International Journal for Parasitology 36 (2006) 79–86
and solving Eq. (1) for t. To estimate confidence intervals for t, the defined distributions of N8, a, b and c were sampled 10,000 times across the distributions calculated from their likelihood profile. This generated 10,000 values for t, and the 250th and 9750th ranked samples were taken as the lower and upper 95% confidence limits, respectively. Statistical differences between times were based on the rank percentile that the mean time to excrete the relevant percentage burden occurred in the comparison group. Estimates of the mean egg excretion per parasite were made by calculating the overall egg production (NN) and dividing this by the worm burden isolated at 35 dpi for each host species. Confidence intervals were estimated by Monte-Carlo techniques using the defined distribution of worm burden and biotic potential calculated from the likelihood profile. The probability of differences between egg productions of worms obtained from different host species was calculated in a similar manner. 3. Results Worms established in the intestines of all host species, but infection frequency and worm burden varied significantly (Table 1). Worms were found in foxes and raccoon dogs at all Table 1 Worm burdens of Echinococcus multilocularis found at necropsy in cats, dogs, raccoon dogs, and foxes (nZ5) experimentally inoculated with 20,000 protoscolices
Individual worm burdens at 35 dpi
Mean worm burden Lower 95% CI Upper 95% CI Individual worm burdens at 63 dpi
Mean worm burden Lower 95% CI Upper 95% CI Individual worm burdens at 90 dpi
Mean worm burden Lower 95% CI Upper 95% CI
Fox
Raccoon dogs
Dog
Cat
16,580 16,244 11,732 19,704 19,700 16,792a 13,877 20,000 720 100 380 399 55 331f 157 889 63 10 419 160 17 134f 49 579
13,700 5,910 2,020 5,040 12,980 7930b 4365 16,703 3310 2390 7230 1180 1956 3,213d,f 1,820 6,476 140 110 20 28 35 67f,g 33 159
3238 3330 1925 2905 930 2466c 1609 4053 85 85 193 3 4130 899f 232 10,208 0 60 1,785 5,420 550 1,563d 332 20,000
1250 25 1690 0 244 642 144 11,748 120 0 5 16 0 28e,f 5 1,429 285 2 0 0 0 57 4 Indeterminate
dpi, days post inoculation; CI, confidence interval. a Significantly higher than in racoon dogs, dogs and cats at this time point. b Significantly higher than dogs and cats at this time point. c Significantly higher than cats at this time point. d Significantly higher than in the three other species at this time point. e Significantly lower than the three other species at this time point. f Significantly less than at 35 dpi in the same species. g Significantly less than at 63 dpi in the same species.
times, and in 14 of 15 dogs (one was negative at 90 dpi). In cats, worms were detected in four of five at 35 dpi, three animals at 63 dpi and two animals at 90 dpi. No clinical signs, which could be assigned to the parasite infection, were observed in any animals during the experiment. Individual worm burdens are summarised in Table 1. At 35 dpi, mean worm burdens representing 84, 39, 12, and 3.2% of protoscolices inoculated were counted in foxes, raccoon dogs, dogs and cats, respectively. Significant differences between worm burdens of different species at different time points are indicated in Table 1. There were no significant differences between egg counts from female and male animals and so data was pooled for further analysis. A number of alternative models were fitted to the egg count data. These included the sigmoidal models: the Verhulst growth curve and the models described by Eqs. (1)– (3). The best fit was the Richard’s growth curve with four parameters (Eqs. (2) and (3)) which gave a significant fit to the data from dogs, racoon dogs and foxes (P!0.005, P!0.01, P!0.025). Simple polynomial or exponential growth model models tended to gave a poorer fit (data not shown). The egg excretion dynamics, based on the results of the Richard’s model for all four species is reported in Table 2. There were no significant differences between the total biotic potential between foxes, racoon dogs and dogs. Figs. 1–3 display the model fit, the 95% confidence intervals together with the observed data for foxes, dogs and raccoon dogs The cats had much lower total egg output and egg excretion was sporadic. In total, only four eggs were detected in a single faecal sample 28 dpi whilst a single egg was detected in three other samples from different animals at 34–54 dpi. Hence, model parameters could not be derived with any accuracy for this species. The egg excretion, determined as epg x mean faecal mass, varied significantly over time and between the host species (Figs. 1–3). The average daily faecal masses were determined as 14.9 g for the foxes, 13.1 g for the raccoon dogs, 68.9 g for the dogs and 10.7 g for the cats. The model suggested that the maximum likelihood estimates (MLE) for the time to excretion of 50, 75 and 95% of the biotic potential was shortest in the fox and longest in the dog. However, the confidence intervals show that these differences were not significant (Table 3). The small numbers of eggs produced in infected cats were eliminated early (Table 3). The 95% CIs of the effective patent period, defined as the time taken for 95% of the eggs to be excreted Table 2 Maximum likelihood estimates and their 95% confidence intervals (CIs) of the biotic potential (total eggs excreted) following experimental inoculation with 20,000 protoscolices of Echinococcus multilocularis
Fox Raccoon dog Dog Cat
Biotic potential
CIs
346,473 335,361 279,910 573
230,420–581,130 206,490–578,819 183,559–463,719 Indeterminatea
The biotic potential in the cat was significantly lower than the other three species. a Gave a poor fit to the model.
Number of eggs per animal
Number of eggs per animal
C.M.O. Kapel et al. / International Journal for Parasitology 36 (2006) 79–86
A
40000 30000 20000 10000 0
0
10
20
30
40
50
60
70
80
A
40000 30000 20000 10000 0
0
0.80
A405nm
A405nm
0.40 0.20 0
10
20
30
40
50
60
70
80
40
50
60
70
80
10
20 30 40 50 60 70 Days after experimental infection
90
B
0.40
0.00
90
Fig. 1. (A) Dynamics of Echinococcus multilocularis egg excretion in faeces following experimental inoculation of 20,000 protoscolices in foxes. Observed data (C) (mean of recorded data at respective time points). Solid line is best fit to logistic model (–) and 95% confidence intervals (/) (P!0.01). Observed data indicated as (B) had evidence of lack of fit to the model. (B) Observed dynamics of detected coproantigen levels measured by ELISA in the faeces (GSD) of foxes.
was 17–44 days for foxes, 22–47 days for raccoon dogs and 22–93 days for dogs (Table 3). An estimate of 13 days was made for cats, but CIs could not be calculated for this species. The model suggested that the highest egg output occurred in Number of eggs per animal
30
0.20 0
Days after experimental infection
A
40000 30000 20000 10000 0 0
10
20
30
40
50
60
70
80
90
B
0.80 0.60 A 405 nm
20
0.60
0.60
0.40 0.20 0.00
10
90
B
0.80
0.00
83
0
10
20 30 40 50 60 70 Days after experimental infection
80
90
Fig. 2. (A) Dynamics of Echinococcus multilocularis egg excretion in faeces following experimental inoculation of 20,000 protoscolices in raccoon dogs. Observed data (C) (mean of recorded data at respective time points). Solid line is best fit to logistic model (–) and 95% confidence intervals (/) (P!0.025). Observed data indicated as (B) had evidence of lack of fit to the model. (B) Observed dynamics of detected coproantigen levels measured by ELISA in the faeces (GSD) of racoon dogs.
80
90
Fig. 3. (A) Dynamics of Echinococcus multilocularis egg excretion in faeces following experimental inoculation of 20,000 protoscolices in dogs. Observed data (C) (mean of recorded data at respective time points). Solid line is best fit to logistic model (–) and 95% confidence intervals (/) (p!0.005). (B) Observed dynamics of detected coproantigen levels measured by ELISA in the faeces (GSD) of dogs.
foxes 37–42 dpi, in raccoon dogs 38–43 dpi and in dogs 43– 45 dpi (Figs. 1–3). The highest daily egg excretion determined were 98,638 for a fox (40 dpi), 88,032 for a raccoon dog (45 dpi), 112,996 eggs for a dog (60 dpi) and 856 for a cat (28 dpi). The dynamics of coproantigen excretion for foxes, raccoon dogs and dogs are shown in Figs. 1–3. In all four animal species, specific reaction was found only sporadically within the first 10 dpi. Although specific coproantigen reactions were observed for cats from 21 dpi until the end of the experiment, around 50% of the samples gave high non-specific reactions, which made the results uninterpretable (data not shown). During 10–15 dpi, 75% of samples from foxes, 94% of samples from raccoon dogs and 31% samples from dogs were coproantigen positive. During 20–28 dpi, 100% of samples from both foxes and racoon dogs were positive and 86% of dog samples were positive. During the first part of patency (29–63 dpi), 82, 79 and 91% of samples from foxes, racoon dogs and dogs were coproantigen positive. In second part of patency, the figures were 51, 17 and 52%, respectively. Table 3 Time taken in days (G95% confidence intervals (CIs)) to expel 50, 75 and 95% of the total eggs excreted in cats, dogs, raccoon dogs and foxes experimentally inoculated with 20,000 protoscolices of Echinococcus multilocularis Proportion of total egg mass
50%
75%
95%
Fox Raccoon dog Dog Cat
9.9 (2.7–22.7) 15.9 (9.4–28.8) 18.8 (3.2–55.6) 6.7a
13.4 (5.7–27.1) 18.8 (12.0–32.5) 23.8 (7.1–63.4) 10.0a
26.8 (17.0–44.2) 30.3 (22.1–46.8) 43.0 (21.9–93.1) 13.0a
a
Indeterminate CIs.
84
C.M.O. Kapel et al. / International Journal for Parasitology 36 (2006) 79–86
Table 4 Infectivity in mice of Echinococcus multilocularis eggs isolated from either faeces or from gravid worms of experimentally infected dogs, raccoon dogs and foxes (number of isolated eggs from faeces of cats was too low for testing) Hosts
Source of eggs
Age of infection (dpi)
Dose per mouse
No. mice developing cysts
Foxes Raccoon dogs Dogs Foxes Raccoon dogs Cats
Faeces Faeces Faeces Worms Worms Worms
35 35 35 63 63 63
2000 2000 2000 2000 2000 2800
5 5 5 5 5 0
In each case, five experimental mice were infected, except with eggs from worms in cats, where only two mice were infected. dpi, days post inoculation.
The bioassay evaluating the infectivity of excreted eggs (Table 4) revealed that all mice inoculated with eggs isolated 35 dpi from the faeces of foxes, raccoon dogs and dogs developed metacestodes with fully developed protoscolices 6 months later. The low epg in cats did not allow for inoculation of mice with eggs from faeces. Eggs, which as an alternative were isolated directly from gravid proglottids of raccoon dogs and foxes at 63 dpi, resulted in fully developed metacestodes whereas eggs from proglottids of cats did not establish. The estimated mean number of eggs excreted per parasite for foxes was 27 (95% CIs 17–44), for raccoon dogs it was 42 (19–85), for dogs it was 114 (63–215), whilst for cats it was only 0.76 (0.02–22). Worms from dogs produced significantly more eggs than those from foxes (P!0.001) and raccoon dogs (P!0.01). There was no significant difference in the mean number of eggs produced per worm between foxes and raccoon dogs. 4. Discussion The experimental infections described provide important information on the biotic potential of E. multilocularis, the establishment of the infection, the dynamics of egg excretion, and the life-span of the infection. It is important to note that the data presented may not be applicable to other definitive hosts such as the arctic fox (A. lagopus) which may have different dynamics of infection. Foxes have a higher initial worm establishment with a shorter parasite life expectancy compared with the racoon dog and dog. This results in an earlier peak of egg production with a shorter period of high egg production, (Table 1, Figs. 1–3). In naturally infected foxes, worm burdens were found to be both highly over-dispersed and to be significantly higher in juvenile foxes compared with adults (Hofer et al., 2000; Raoul et al., 2001; Yimam et al., 2002). Because the dynamics of the infection in foxes appears to be characterised by a rapid loss of the worm burden, despite high initial numbers, foxes with low worm numbers are more likely to be sampled in a given study. However other factors such as different predation habits of the different age groups of foxes, host immunity or genetic susceptibility could also contribute to this phenomenon. Protoscolex numbers in naturally infected voles have also
been shown to be over-dispersed (Stieger et al., 2002). All the experimental animals were young. Hence, it is possible that the biotic potential calculated overestimates the true biotic potential in naturally exposed and possibly less susceptible adult animals. However, it is the highly infected individuals carrying thousands of fertile worms that are responsible for most of the egg contamination. Furthermore, Budke et al. (2005) reported there were no age-related changes in abundance in naturally infected dogs. There are few other data concerning the egg production and dynamics of worm burdens during infection with E. multilocularis. Nonaka et al. (1996) reported that in foxes inoculated with about 150,000 protoscolices, eggs were regularly detected in faeces from 29–60 dpi and sporadically until 84 dpi. Nevertheless, 11 and 91 parasites were detected in two animals 125 dpi. Lukashenko (1975) recovered 70,000 worms (from an infecting dose of 100,000 protoscolices) from dogs 3–4 weeks after inoculation, but found only 528 worms after 3–4 months and just 27 after 6 months. Interestingly, in the study of Vogel (1957) worm burdens ranging from less than 10 specimens to several thousand were found in dogs 42–97 dpi (with 0–55% gravid worms). However, these experiments were performed with different isolates of E. multilocularis and no information on protoscolex numbers was given. In the present study, significant numbers of worms were found at 90 dpi (Table 1) although very few eggs were found in the faeces at this time (Fig. 3). Cats appear to be poor definitive hosts with just 3.2% of the infecting dose establishing (Table 1). Likewise, Lukashenko (1975) recovered 7% of an infecting dose of 50,000 protoscolices 3–4 weeks after infection and just 0.3% after 3 months. Jenkins and Romig (2000, 2003) and Thompson et al. (2003) also recovered only 6.9, 3 and 12.7%, respectively, of the infecting dose of between 10,000 and 22,600 protoscolices between 21 and 25 dpi. Reports of naturally infected cats are characterised by low infection intensity. Petavy et al. (2000) found three of 81 necropsied cats naturally infected but with a mean intensity of just nine worms per cat. Deplazes et al. (1999) described one of 130 stray cats infected with just five parasites. A variety of mathematical equations were assessed to model the egg production dynamics and Eq. (3) gave the most satisfactory fit for the egg count data for foxes, racoon dogs and dogs. The poor fit for cats was probably due to the lack of data resulting from very low egg production. The model fit is consistent with egg production starting slowly at the start of patency, accelerating as the worms mature and then slowing as the worms become senescent. In the fox and raccoon dog, there was some evidence of lack of fit due to two and three data points, respectively (Figs. 1 and 2). Visual inspection suggests that these model outliers could be due to a second, but lower, peak in egg production. For E. multilocularis it has never been convincingly shown that sequential proglottids even develop to gravidity. The relatively short life span of the worms in foxes in this experiment supports the hypothesis that sequential proglottidisation did not contribute significantly to total egg production in this study. However, the model outliers in the fox and racoon
C.M.O. Kapel et al. / International Journal for Parasitology 36 (2006) 79–86
dog (Figs. 1 and 2) could indicate a biphasic response and hence a second wave of proglottid shedding. In dogs, the much higher egg production per worm, and the longer survival of the worms suggest that more than one proglottid per worm may contribute to egg counts but there was little evidence of a biphasic response in the model. Further studies should focus on the developmental processes of growth, segmentation, proglottidisation and maturation in adult Echinococcus to develop a better understanding of its reproductive biology (Thompson et al., unpublished data). The high biotic potential of infections from raccoon dogs demonstrates that the parasite in this host is of particular concern as the range of this species is expanding in Europe. Likewise, the high biotic potential in dogs demonstrates that these hosts could make a significant contribution to the transmission cycle. In some regions such as China, Alaska or Central Europe, dogs may have high prevalence rates (ranging from 7 to O20%) (Schantz et al., 1995; Gottstein et al., 2001; Budke et al., 2005), but even in the general dog population of Switzerland where there is a low prevalence of just 0.3%, the lifetime incidence of infection in dogs could be 10% (Deplazes et al., 2004). In addition, the fact that dogs can carry contamination with eggs on their fur and their close proximity to humans potentiate the zoonotic importance of this host species. Epidemiological studies in China and in Europe have suggested dog ownership and/or close contact with dogs are important risk factors for human alveolar echinococcosis (Craig et al., 2000; Wang et al., 2001; Kern et al., 2004). Descriptions of patent infections in naturally infected cats have been described in the single infected cat reported by Deplazes et al. (1999) and in one cat of 33 examined by Gottstein et al. (2001). No studies of naturally infected cats have demonstrated parasite abundance levels or levels of egg excretion that could be considered important in epidemiological terms. Nevertheless, because the experimental design of the present study necessarily involved relatively few animals, the results cannot absolutely rule out the cat as a possible host for parasite transmission. Although a similar biotic potential was recorded in three host species (Tables 1 and 2) differences in worm establishment rate resulted in a higher biotic potential per parasite in dogs (114) compared with foxes (27) and raccoon dogs (42). The high worm fecundity in dogs could result from a crowding effect caused by the high burdens in foxes compared with dogs. In a limited experiment, not analysed statistically, Heath and Lawrence (1991) suggested that three dogs experimentally inoculated with 88,000 protoscolices of E. granulosus reached patency earlier with a more rapid increase in faecal egg production compared with inoculations of 44,000 protoscolices. Alternatively, the immunological response to E. multilocularis may be more efficient in foxes resulting in more rapid expulsion of worms. The estimated numbers of eggs excreted per worm in faeces is low in comparison with the 200 eggs per gravid proglottid reported by Vogel (1957). This difference may be because a large number of non-viable eggs disintegrate in the intestine of the definitive hosts: the egg viability, measured in vitro by the resistance of onchospheres to sodium
85
hypochlorite, was only about 46% in eggs isolated from gravid proglottids (Deplazes et al., 2005). This compares with more than 99% for eggs isolated from faeces of the same hosts. It is also likely that faecal egg counts were underestimated. In a limited comparative study with the sieving technique (Mathis et al., 1996), the method used in this study had an egg recovery rate of approximately 70% (data not shown). In foxes, the peak levels of coproantigen production were a few days earlier than the peak egg production (Fig. 2) with levels declining from 35–45 dpi. This dynamic is similar to that recorded by Nonaka et al. (1996) and appears to mirror the decline in egg production and numbers of worms recovered at necropsy. Likewise, Nonaka et al. (1996) recovered very small numbers of adult worms at the end of their experimental infections. The peak coproantigen production in foxes could correspond to the most metabolically active stage of the infection with high worm burdens just before the onset of patency and during the main elimination of proglottids and worms. Thus, the coproantigen test will have a higher likelihood of detecting infection in the late prepatent and early patent phases of infection. This is important, as the early patent period in foxes appears to be the most significant in terms of transmission. Likewise, the coproantigen production in raccoon dogs showed high levels in the late prepatent period, but also a second peak at about 60 dpi (Fig. 2). Whilst the first peak can be interpreted in a similar manner to that in the foxes, the second peak may be an expression of loss of worms following loss of the terminal egg containing proglottids. In the dog (Fig. 3), the early peak in coproantigen production was absent and maximum levels were seen around 50–60 dpi, corresponding to the observed peak in epg and the extended persistence of the intestinal worms. For foxes, raccoon dogs and dogs, coproantigen levels were detected in a high proportion of samples during the second half of pre-patency and the first 30 days of patency confirming the usefulness of this tool in identifying infected animals. Although coproantigen was detected in cats (data not shown), this was at low and intermittent levels and hence no conclusions can be drawn in this species. Acknowledgements The Danish Veterinary and Agricultural Research Council and the Danish National Research Foundation are thanked for substantial financial support to the present study. The competent staff at the research facility “Lindholm” of the Danish Institute for Food and Veterinary Research is thanked for hard work, flexibility and logistical support. References Asikainen, J., Mustonen, A., Hyva¨rinen, H., Nieminen, P., 2003. Seasonal reproduction endocrine profile of the raccoon dog (Nyctereutes procyonoides)—effects of melatonin and food deprivation. J. Exp. Zoo. Part A: Comp. Exp. Biol. 299, 180–187. Budke, C.M., Craig, P.S., Jiamin, Q., Torgerson, P.R., 2005. Modeling the transmission of Echinococcus multilocularis and Echinococcus granulosus in dogs for a high endemic region of the Tibetan plateau. Int. J. Parasitol. 35, 163–170.
86
C.M.O. Kapel et al. / International Journal for Parasitology 36 (2006) 79–86
Craig, P.S., Giraudoux, P., Shi, D., Bartholomat, B., Barnish, G., Delattre, P., Quere, J.P., Harraga, S., Bao, G., Wang, Y., Lu, F., Ito, A., Vuitton, D.A., 2000. An epidemiological and ecological study of human alveolar echinococcosis transmission in south Gansu. China. Act. Trop. 77, 167–177. Deplazes, P., Alther, P., Tanner, I., Thompson, R.C.A., Eckert, J., 1999. Echinococcus multilocularis coproantigen detection by enzyme-linked immunosorbent assay in fox, dog, and cat populations. J. Parasitol. 85, 115–121. Deplazes, P., Hegglin, D., Gloor, S., Romig, T., 2004. Wilderness in the city: the urbanisation of Echinococcus multilocularis. Trends Parasitol. 20, 77–84. Deplazes, P., Grimm, F., Sydler, T., Tanner, I., Kapel, C.M.O., 2005. Experimental alveolar echinococcosis in pigs, lesion development and serological follow up. Vet. Parasitol. 130, 213–222. Eckert, J., Deplazes, P., 2004. Biological, epidemiological and clinical aspects of echinococcosis, a zoonosis of increasing concern. Clin. Microbiol. Rev. 17, 107–135. Efron, B., Tibshirani, R.J., 1993. An Introduction to the Bootstrap. Chapman & Hall, New York, NY. Gottstein, B., Saucy, F., Deplazes, P., Reichen, J., Demierre, G., Busato, A., Zuercher, C., Pugin, P., 2001. Is high prevalence of Echinococcus multilocularis in wild and domestic animals associated with disease incidence in humans? Emerg. Infect. Dis. 7, 408–412. Heath, D.D., Lawrence, S.B., 1991. Daily egg production of dogs infected with Echinococcus granulosus. Arch. Int. Hidatid. 30, 321–328. Hilbourn, R., Mangel, M., 1997. The Ecological Detective: Confronting Models with Data Monographs in Population Biology, 28. Princeton University Press, New Jersey. Hofer, S., Gloor, S., Mu¨ller, U., Mathis, A., Hegglin, D., Deplazes, P., 2000. High prevalence of Echinococcus multilocularis in urban red foxes (Vulpes vulpes) and voles (Arvicola terrestris) in the city of Zurich, Switzerland. Parasitology 120, 135–142. Jenkins, D.J., Romig, T., 2000. Eficacy of DroncitR spot-on (praziquantel) 4% w/v against immature and mature Echinococcus multilocularis in cats. Int. J. Parasitol. 30, 959–962. Jenkins, D.J., Romig, T., 2003. Milbemycin oxime in a new formulation combined with praziquantel, does not reduce the efficacy of praziquantel against Echinococcus multilocularis in cats. J. Helminthol. 77, 367–370. Kern, P., Ammon, A., Kron, M., Sinn, G., Sander, S., Petersen, L.R., Gaus, W., Kern, P., 2004. Risk factors for alveolar echinococcosis in human. Emerg. Infect. Dis. 10, 2088–2093. Lukashenko, N.P., 1975. Alveokokkoz, Meditsina, Moskva 1975. 327 pp. (In Russian). Mathis, A., Deplazes, P., Eckert, J., 1996. Improved test system for PCR-based specific detection of Echinococcus multilocularis eggs. J. Helminthol. 70, 219–222. Nonaka, N., Iida, M., Yagi, K., Ito, T., Ooi, H.K., Oku, Y., Kamiya, M., 1996. Time course of coproantigen excretion in Echinococcus multilocularis
infections in foxes and an alternative definitive host, golden hamsters. Int. J. Parasitol. 26, 1271–1278. Petavy, A.F., Tenora, F., Deblock, S., Sergent, V., 2000. Echinococcus multilocularis in domestic cats in France. A potential risk factor for alveolar hydatid disease contamination in humans. Vet. Parasitol. 87, 151–156. Pruˆsaite`, J., Mazeikyte, R., Pauza, D., 1988. Fauna of Lithuania (Lietuvos Fauna). Mokslav, Vilnius, p. 186–218. Raoul, F., Deplazes, P., Nonaka, N., Piarroux, R., Vuitton, D.A., Giraudoux, P., 2001. Assessment of the epidemiological status of Echinococcus multilocularis in foxes in France using ELISA coprotests on fox faeces collected in the field. Int. J. Parasitol. 31, 1579–1588. Richards, F.J., 1959. A flexible growth function for empirical use. J. Exp. Bot. 10, 290–300. Roepstorff, A., Nansen, P., 1998. The Epidemiology. Diagnosis and Control of Helminth Parasites of Swine FAO Animal Health Manual No. 3. FAO, Rome, Italy. Schantz, P.M., Chai, J., Craig, P.A., Eckert, J., Jenkins, D.J., Macpherson, C.N.L., Thakur, A., 1995. Epidemiology and control of hydatid disease. In: Thompson, R.C.A., Lymbery, A.J. (Eds.), Echinococcus and hydatid disease. CAB International, Wallingford, UK, pp. 233–331. Stieger, C., Hegglin, D., Schwarzenbach, G., Mathis, A., Deplazes, P., 2002. Spatial and temporal aspects of urban transmission of Echinococcus multilocularis. Parasitology 124, 631–640. Thiess, A., Schuster, R., Nockler, K., Mix, H., 2001. Helminth findings in indigenous raccoon dogs Nyctereutes procyonoides (Gray, 1843).. Berl. Mu¨nch. Tierarztl. Wschr. [German] 114, 176–273. Thompson, R.C.A., Deplazes, P., Eckert, J., 2003. Observations on the development of Echinococcus multilocularis in cats. J. Parasitol. 89, 1086–1088. Torgerson, P.R., Shaikenov, B.S., Rysmukhambetova, A.T., Abdybekova, A.M., Usenbayev, A.E., Baitursinov, K.K., 2003. Modelling the transmission dynamics of Echinococcus granulosus in dogs in rural Kazakhstan. Parasitology 126, 417–424. Venzon, D.J., Moolgavkar, S.H., 1988. A method for computing profilelikelihood-based confidence intervals. Appl. Stat. 37, 87–94. Vogel, H., 1957. Echinococcus multilocularis in South Germany. I. The tapeworm stage of strains from humans and animals. Z. Tropenmed. Parasitol. 8, 404–454. Vuitton, D.A., Zhou, H., Bresson-Hadni, S., Wang, Q., Piarroux, M., Raoul, F., Giraudoux, P., 2003. Epidemiology of alveola echinococcosis with particular reference to China and Europe. Parasitology 127, S87–S107. Wang, Q., Qiu, J., Schantz, P., He, J.G., Ito, A., Liu, F., 2001. Investigation of risk factors for development of human hydatidosis among households raising livestock in Tibetan areas of western Sichuan Province. Chin. J. Parasitol. Parasit. Dis. 19, 93–96. Yimam, A.E., Nonaka, N., Oku, Y., Kamiya, M., 2002. Prevalence and intensity of Echinococcus multilocularis in red foxes (Vulpes vulpes schrencki) and raccoon dogs (Nyctereutes procyonoides albus) in Otaru City, Hokkaido, Japan. Jpn. J. Vet. Res. 49, 287–296.