INTRODUCTION
Animal health management has been defined as the promotion of health, improvement of productivity and prevention of disease within the economic framework of farm owner and industry, while recognizing animal welfare, food safety and security, public health and environmental sustainability (Leblanc et al. Reference Leblanc, Lissemore, Kelton, Duffield and Leslie2006). Prevention and vaccination campaigns are resulting in successful local and regional elimination of several infectious diseases of cattle such as foot and mouth disease (Sutmoller et al. Reference Sutmoller, Barteling, Olascoaga and Sumption2003), infectious bovine rhinothracheitis (Nardelli et al. Reference Nardelli, Farina, Lucchini, Valorz, Moresco, Dal Zotto and Costanzi2008) and bovine viral diarrhoea (Lindberg et al. Reference Lindberg, Brownlie, Gunn, Houe, Moennig, Saatkamp, Sandvik and Valle2006). In contrast, elimination of endemic parasitic diseases by vaccination has not occurred and is probably unrealistic. In fact, increases in helminth-associated disease frequency and intensity have been reported within the European ruminant sector in recent years (Morgan et al. Reference Morgan, Charlier, Hendrickx, Biggeri, Catalan, von Samson-Himmelstjerna, Demeler, Müller, van Dijk, Kenyon, Skuce, Höglund, O'Kiely, Van Ranst, de Waal, Rinaldi, Cringoli, Torgerson, Hertzberg, Wolstenholme and Vercruysse2013). This includes a substantial increase in laboratory diagnoses of ovine and bovine fasciolosis in the UK, where central collation of such data allows fair temporal comparisons (van Dijk et al. Reference van Dijk, Sargison, Kenyon and Skuce2010). Given this apparent trend, and the significant economic and welfare burden of fasciolosis in cattle (Schweizer et al. Reference Schweizer, Braun, Deplazes and Torgerson2005; Charlier et al. Reference Charlier, Sanders and Vercruysse2009a ), it is timely to re-visit understanding of this disease. Advances in the last decade in diagnosis underpin new knowledge and understanding of impacts on production, including the effect of the host immune response to co-infections and the prediction of disease. In each section, we address these areas and identify remaining knowledge gaps. We conclude with a discussion of research priorities to promote better management of bovine fasciolosis within an economic framework.
BOVINE FASCIOLOSIS: A RESURGENT DISEASE IN EUROPE?
Regional increases in laboratory diagnoses of fasciolosis in cattle and sheep (van Dijk et al. Reference van Dijk, Sargison, Kenyon and Skuce2010; Fairweather, Reference Fairweather2011), and anecdotal reports from veterinary practitioners of ever-increasing challenges from this disease, raise concerns that risks of infection are indeed increasing. Altered disease patterns could be detected by changes in the prevalence, intensity or distribution of Fasciola hepatica (Mas-Coma et al. Reference Mas-Coma, Valero and Bargues2008). The distribution can be affected (1) temporally, causing shifts in seasonality or (2) spatially, causing shifts in the geographic range of occurrence. For fasciolosis, increases in prevalence and shifts in the geographic range have been reported in the last decade, primarily in the UK. Fasciolosis was reported in previously unaffected parts of Scotland and East Anglia (Pritchard et al. Reference Pritchard, Forbes, Williams, Salimi-Bejestani and Daniel2005; Kenyon et al. Reference Kenyon, Sargison, Skuce and Jackson2009). In England, prevalence of infection in dairy herds was 48% in 2003 compared with 72% in 2006 (McCann et al. Reference McCann, Baylis and Williams2010a ). In these cases, the authors linked increases in prevalence to milder winter temperatures and increased rainfall.
The use of different diagnostic methods or sampling in different areas hampers comparison of the results from prevalence surveys across space and time. Nonetheless, in some enzootic regions, repeated surveys with comparable methodology were carried out over the last decade (Table 1). In north-western Spain, Arias et al. (Reference Arias, Piñeiro, Hillyer, Suárez, Francisco, Cortiñas, Díez-Baños and Morrondo2010) found a sero-prevalence of 65% and concluded that despite regular fasciolicide treatment, the control of fasciolosis had not improved in this area compared with a survey carried out 10 years earlier (Sánchez-Andrade, Reference Sánchez-Andrade, Paz-Silva, Suárez, Panadero, Díez-Baños and Morrondo2000). In Switzerland, Rapsch et al. (Reference Rapsch, Schweizer, Grimm, Kohler, Bauer, Deplazes, Braun and Torgerson2006) reported a higher prevalence (18%) than those previously reported (8–15%) in slaughtered cattle. However, the authors concluded that this was mainly due to the higher sensitivity of the detection techniques in their survey compared with the previous ones. In Belgium, no decreases in prevalence were observed between 2006–2008 (Bennema et al. Reference Bennema, Ducheyne, Vercruysse, Claerebout, Hendrickx and Charlier2011) and 2009–2011 (Charlier et al. Reference Charlier, Meyns, Soenen and Vercruysse2013) during which an active monitoring campaign increased awareness of the disease in this region.
Table 1. Comparable surveys of F. hepatica in cattle in three different parts of Europe. Only those surveys in which the sampled population and/or the test used were very similar were included. Prev. = prevalence; Ab-det. = antibody detection
Studies in other regions do not allow trends to be assessed over time but show an overall moderate to high prevalence of F. hepatica in different regions from southern to sub-Scandinavian Europe (Cringoli et al. Reference Cringoli, Rinaldi, Veneziano, Capelli and Malone2002; Conceição et al. Reference Conceição, Durao, Costa, Castro, Louza and Costa2004; Kuerpick et al. Reference Kuerpick, Conraths, Staubach, Schnieder and Strube2013a ). For example, studies in Germany (Pfister and Koch, Reference Pfister and Koch2004; Kuerpick et al. Reference Kuerpick, Conraths, Staubach, Schnieder and Strube2013a ) and Austria (more specifically Tirol/Carinthia) (Matt et al. Reference Matt, Schopf and Mader2007; Duscher et al. Reference Duscher, Duscher, Hofer, Tichy, Prosl and Joachim2011) report overall herd-level prevalences of 24 and > 70%, respectively and show that herd-level prevalence reaches very high values (up to 97%) in alpine upland farms. In northern Europe, the parasite seems not to be widely distributed, with a reported herd-level prevalence of 7% in south-central Sweden (Höglund et al. Reference Höglund, Dahlstrom, Engstrom, Hessle, Jakubek, Schnieder, Strube and Sollenberg2010).
The role of climatic factors (milder winter temperatures and higher amounts of rainfall) in increased prevalence of F. hepatica seems obvious. However, other, more complex environmental changes could very plausibly affect observed prevalence and disease risk. In England, local increases in the prevalence of fasciolosis in cattle in East Anglia were associated with increased use of sheep to graze environmentally sensitive areas (Pritchard et al. Reference Pritchard, Forbes, Williams, Salimi-Bejestani and Daniel2005). These areas were generally on wet land, resulting from the reversal of previous land drainage schemes in support of conservation and watershed management, while the sheep were imported from areas of the UK where F. hepatica was enzootic. In contrast, reversal of pasture drainage measures for nature conservation in northern Germany did not result in detectable changes in fasciolosis in cattle over a 3-year period (Kemper and Henze, Reference Kemper and Henze2009). This suggests that the local drivers of transmission and disease risk are complex and to some extent site-specific, and could be important modifiers of predicted macro-scale trends driven by global climate change (Fox et al. Reference Fox, White, McClean, Marion, Evans and Hutchings2011). This is borne out by statistical models using bulk milk tank antibody levels as a measure of exposure to F. hepatica, which are able to explain around 20% of variance when only geographic or climatic factors are considered (Kuerpick et al. Reference Kuerpick, Conraths, Staubach, Schnieder and Strube2013a ), rising to 85% when more detailed local factors are also included (Charlier et al. Reference Charlier, Bennema, Caron, Counotte, Ducheyne, Hendrickx and Vercruysse2011).
Considering the published prevalence surveys of F. hepatica in European cattle herds over the last decade, it is clear that F. hepatica probably affects cattle in every EU member state. Evidence of increasing prevalence of F. hepatica is only reported in the UK. Despite the lack of recent published information in some countries with important cattle industries (for example France and countries in Eastern Europe), it is reasonable to say that in other regions throughout Europe, the prevalence is not decreasing despite the greater attention given to control in recent years. In addition, there are cases in which local environmental change appears to have driven disease emergence, while macro-climatic change is predicted to favour rather than hinder transmission. Certainly, there is little basis on which to allay the significant concerns over this disease among animal health professionals in Europe.
IMPROVING DIAGNOSIS
The pursuit of the perfect diagnostic test
The classical cornerstone for the diagnosis of fasciolosis in cattle is the microscopic detection of F. hepatica eggs in feces. More recently important advances have been made in alternative methods based on detection of F. hepatica-specific antibodies in serum or milk and the detection of F. hepatica-specific antigens or DNA in the feces.
Detection of eggs is mostly based on sedimentation of eggs, which may be followed by their flotation. Numerous methods have been described in literature and in practice each laboratory has adopted its own variant. In general, the trained eye can easily recognize F. hepatica eggs using low magnification conventional microscopy and discriminate between them and others such as Paramphistomum spp. eggs. Although the specificity of these methods is high and the detection of eggs indicates current infection (instead of exposure as with antibody detection), a disadvantage of detection of eggs in feces, particularly in cattle, is the low sensitivity (30–70%, depending on the method and study area). However, recent studies have demonstrated that the sensitivity of egg detection methods is largely driven by the volume of feces that is analysed (Conceição et al. Reference Conceição, Durao, Costa and da Costa2002; Rapsch et al. Reference Rapsch, Schweizer, Grimm, Kohler, Bauer, Deplazes, Braun and Torgerson2006; Charlier et al. Reference Charlier, De Meulemeester, Claerebout, Williams and Vercruysse2008) and repeated testing or analysing ⩾30 g of feces can increase the sensitivity up to 90% (Rapsch et al. Reference Rapsch, Schweizer, Grimm, Kohler, Bauer, Deplazes, Braun and Torgerson2006).
Detection of F. hepatica-specific antibodies, mostly by ELISA, has been proposed as a more sensitive method. There are several ELISAs that have been well described in the literature. The first, referred to as ES ELISA, uses the complete excretory secretory (ES) products of F. hepatica. ES products are relatively easy to produce and as a consequence many laboratories apply their own in-house version of the ELISA. Recently, a more standardized ELISA kit has become available (Svanovir® F. hepatica-Ab, Svanova, Uppsala, Sweden). Although cross-reactions of the F. hepatica ES products with antibodies against other helminths such as Dictyocaulus viviparus (Cornelissen et al. Reference Cornelissen, Gaasenbeek, Boersma, Borgsteede and van Milligen1999) and Dicrocoelium dendriticum (Mezo et al. Reference Mezo, Gonzalez-Warleta and Ubeira2007) are reported, reasonable sensitivities (86–100%) and specificities (83–96%) of ES ELISAs have been observed under different epidemiological settings (Anderson et al. Reference Anderson, Luong, Vo, Bui, Smooker and Spithill1999; Cornelissen et al. Reference Cornelissen, Gaasenbeek, Boersma, Borgsteede and van Milligen1999; Salimi-Bejestani et al. Reference Salimi-Bejestani, McGarry, Felstead, Ortiz, Akca and Williams2005; Charlier et al. Reference Charlier, De Meulemeester, Claerebout, Williams and Vercruysse2008; Kuerpick et al. Reference Kuerpick, Schnieder and Strube2013b ). A second ELISA uses a sub-fraction of the ES products as diagnostic antigen, the so-called ‘f2’ antigen, for which the purification steps are described by Tailliez and Korach (Reference Tailliez and Korach1970). A commercial format (Fasciolosis Verification Test, IDEXX, Hoofddorp, the Netherlands; previous manufacturer: Institut Pourquier) and several evaluations of this kit are available, reporting high sensitivity (88–98%) and specificity (84–98%) (Reichel, Reference Reichel2002; Molloy et al. Reference Molloy, Anderson, Fletcher, Landmann and Knight2005; Rapsch et al. Reference Rapsch, Schweizer, Grimm, Kohler, Bauer, Deplazes, Braun and Torgerson2006; Charlier et al. Reference Charlier, De Meulemeester, Claerebout, Williams and Vercruysse2008; Kuerpick et al. Reference Kuerpick, Schnieder and Strube2013b ). The MM3-ELISA developed by Mezo et al. (Reference Mezo, Gonzalez-Warleta, Castro-Hermida, Muino and Ubeira2010) uses monoclonal antibodies to capture proteins from a 7–40 kDa fraction (mainly cathepsins L1 and L2) from the ES products (Muino et al. Reference Muino, Perteguer, Garate, Martinez-Sernandez, Beltran, Romaris, Mezo, Gonzalez-Warleta and Ubeira2011). A field evaluation reported a sensitivity of 99% and specificity of 100%. A commercial format is available (BIO K 211, Bio-X Diagnostics, Jemelle, Belgium) and further evaluations of this commercial kit are expected in the future. Finally, an ELISA based on recombinant cathepsin L1 is also described, with the major advantage that it does not depend on the availability of living flukes. However, in a direct comparison the sensitivity and specificity were lower than for f2 and ES ELISA, respectively (Kuerpick et al. Reference Kuerpick, Schnieder and Strube2013b ). A recombinant mutant cathepsin L1 ELISA is commercially developed by Ildana biotech (Dublin, Ireland).
Besides increased sensitivity, the major advantage of immunodiagnosis is that it is suitable for high-throughput formats and can be applied on different matrices besides serum. Testing bulk-tank and individual milk samples offers a great advantage in dairy farming as milk samples are collected for other monitoring programmes on a regular basis and there is no need for invasive sampling of the animals (Reichel et al. Reference Reichel, Vanhoff and Baxter2005; Salimi-Bejestani et al. Reference Salimi-Bejestani, Daniel, Cripps, Felstead and Williams2007; Duscher et al. Reference Duscher, Duscher, Hofer, Tichy, Prosl and Joachim2011). A similar approach has been proposed for beef cattle, using muscle transudate (‘meat juice’) collected in the abattoir as matrix for F. hepatica ELISA (Charlier et al. Reference Charlier, De Cat, Forbes and Vercruysse2009b ).
Another technique is the detection of F. hepatica antigens in feces through ELISA (Espino and Finlay, Reference Espino and Finlay1994; Abdel-Rahman et al. Reference Abdel-Rahman, O'Reilly and Malone1998; Mezo et al. Reference Mezo, Gonzalez-Warleta, Carro and Ubeira2004). The test described by Mezo et al. (Reference Mezo, Gonzalez-Warleta, Carro and Ubeira2004), using the MM3-antibody directed towards F. hepatica cathepsins is available commercially (BIO K 201, Bio-X Diagnostics, Jemelle, Belgium). This copro-antigen test is described as an ultra-sensitive and specific test capable of detecting prepatent infections from 4 weeks post-infection onwards and animals infected with as low as 1–2 flukes (Mezo et al. Reference Mezo, Gonzalez-Warleta, Carro and Ubeira2004). Despite recent field evaluations of the commercial test reporting rather low sensitivity (Duscher et al. Reference Duscher, Duscher, Hofer, Tichy, Prosl and Joachim2011; Salem et al. Reference Salem, Chauvin, Braun, Jacquiet and Dorchies2011), the test has been successfully evaluated to detect triclabendazole resistance in sheep by assessing copro-antigen reductions after flukicide treatment (Flanagan et al. Reference Flanagan, Edgar, Gordon, Hanna, Brennan and Fairweather2011; Gordon et al. Reference Gordon, Zadoks, Stevenson, Sargison and Skuce2012).
Finally, DNA-based techniques have so far received relatively little attention in the context of liver fluke diagnosis. This can be justified as the most obvious source of DNA is F. hepatica eggs in the feces and the techniques would thus not offer great advantage above classical microscopical quantification of eggs. However, a recent study in sheep shows that PCR-technology could be useful to detect infections as early as 2 weeks post infection, before serum antibody develops in many infected animals and before eggs appear in feces and this with great sensitivity and specificity (Martinez-Perez et al. Reference Martinez-Perez, Robles-Perez, Rojo-Vazquez and Martinez-Valladares2012). Another promising application is the use of a LAMP (loop mediated isothermal amplification) assay on fecal samples. Ai et al. (Reference Ai, Li, Elsheikha, Hong, Chen, Chen, Li, Cai, Chen and Zhu2010) describe this assay that is even more sensitive than conventional PCR technology. Moreover, DNA amplification is possible under isothermal conditions, quick and visible by the naked eye, thus has potential as a pen-side diagnostic test.
The current paradigm change from diagnosing the infection to its impact
The previous cited work on diagnosis is characterized by a clear progression: the reported tests show, in chronological order, increased sensitivity, specificity and precision compared with previous tests. Yet, this has not (yet) resulted in a major reduction in disease prevalence.
A possible explanation for this is that the tests focus on detecting presence/absence of infection while no clear message on control is provided (Vercruysse and Claerebout, Reference Vercruysse and Claerebout2001). In cattle, fasciolosis is a chronic disease, therefore, there is limited value in knowing the status of presence/absence of infection in a cow or herd or diagnosing infections at 2 weeks vs 8 weeks post-infection; it is more relevant to have some information on intensity of infection and associated production losses to convince farmers that further diagnosis, and treatment, are worth considering.
A study by Charlier et al. (Reference Charlier, De Meulemeester, Claerebout, Williams and Vercruysse2008) showed that while the majority of naturally infected animals had a low fluke burden, signs of liver damage (visible liver lesions or increases in the bile duct enzyme γ-glutamyltransferase) only appeared when >10 flukes were present. Coprology applied on 4 g of feces identified only these highly infected animals, while coprology applied on 10 g of feces also identified animals with lower worm burdens (Charlier et al. Reference Charlier, De Meulemeester, Claerebout, Williams and Vercruysse2008). The copro-antigen ELISA showed good correlation between test results and fluke burden and could also be a valuable aid to control (Mezo et al. Reference Mezo, Gonzalez-Warleta, Carro and Ubeira2004; Charlier et al. Reference Charlier, De Meulemeester, Claerebout, Williams and Vercruysse2008). Antibody detection ELISAs also discriminate to some extent between no, low, moderate and high fluke burdens (Charlier et al. Reference Charlier, De Meulemeester, Claerebout, Williams and Vercruysse2008; Salimi-Bejestani et al. Reference Salimi-Bejestani, Cripps and Williams2008).
Diagnostic test results have been directly correlated with production parameters and with production responses after flukicide treatment (Charlier et al. Reference Charlier, Duchateau, Claerebout, Williams and Vercruysse2007, Reference Charlier, De Cat, Forbes and Vercruysse2009b , Reference Charlier, Hostens, Jacobs, Van Ranst, Duchateau and Vercruysse2012a ; Mezo et al. Reference Mezo, González-Warleta, Castro-Hermida, Muiño and Ubeira2011; Kuerpick et al. Reference Kuerpick, Fiedor, von Samson-Himmelstjerna, Schnieder and Strube2012). Establishing these relationships adds value to the diagnostic test because not only can information be given on the likelihood of the presence of infection, but also on the likely impact it has on animal performance. Moreover, this information can then be used to calculate herd-specific, albeit crude estimates of the direct costs due to F. hepatica (Charlier et al. Reference Charlier, van der Voort, Hogeveen and Vercruysse2012b ) and provide support for decisions on liver fluke control.
It is clear that this new approach of moving from purely detecting infection to an economic diagnosis (assessing the economic impact) is still in its infancy. Production impacts will vary in different epidemiological and farm management settings. Moreover, van der Voort et al. (Reference van der Voort, Charlier, Lauwers, Vercruysse, Van Huylenbroeck and Van Meensel2013) show that the economic impact of production losses will depend on the farm's efficiency of input-output transformation. Previous studies have measured the impact on key production indicators such as milk yield or carcass weight. A much closer collaboration between agricultural economists and parasitologists will be needed to manage parasitic infections within the whole farm economic context (van der Voort et al. Reference van der Voort, Charlier, Lauwers, Vercruysse, Van Huylenbroeck and Van Meensel2013).
IMPACT ON PRODUCTION
Fasciola hepatica infections are reported to affect weight gain, milk production, milk solids content and fertility of cattle (Black and Froyd, Reference Black and Froyd1972; Lopez-Diaz et al. Reference Lopez-Diaz, Carro, Cadorgina, Diez-Banos and Mezo1998; Torgerson and Claxton, Reference Torgerson, Claxton and Dalton1999). Although the general negative impact of F. hepatica on production parameters is well accepted, Dargie (Reference Dargie1987) expressed his concern on the lack of appropriately designed studies to quantify these impacts. These (older) studies are well summarized by Schweizer et al. (Reference Schweizer, Braun, Deplazes and Torgerson2005), who estimated mean losses of a 9% reduction in weight gain in growing cattle, 10% reduction in milk yield, an extension of the service period by 13 days and an increment of 0·75 services per conception.
More recently, a number of additional studies have provided new figures that may provide more representative estimates for modern cattle production systems. Loyacano et al. (Reference Loyacano, Williams, Gurie and DeRosa2002) observed a 6% increase in body weight of beef heifers (N = 372) that were raised under repeated flukicide treatment compared with untreated controls. In abattoir studies Charlier et al. (Reference Charlier, De Cat, Forbes and Vercruysse2009b ) and Sanchez-Vazques and Lewis (Reference Sanchez-Vazques and Lewis2013) related F. hepatica infection status (based on F. hepatica meat juice ELISA and liver condemnation) to carcass parameters, while adjusting for potential confounding factors such as age and herd-clustering effects. Charlier et al. (Reference Charlier, De Cat, Forbes and Vercruysse2009b ) observed a marginal significant mean reduction in warm carcass weight of 0·7% (3·4 kg) in Belgian Blue suckler cows (N = 1450) and no significant effects on conformation score or fat coverage. Sanchez-Vazques and Lewis (Reference Sanchez-Vazques and Lewis2013), using records (N = 328 137) from a 6-year period and different beef breeds assessed a 0·5% reduction in cold carcass weight and a reduction in the price of the carcass by 0·3%. The odds for liver fluke-infected animals receiving a higher conformation or fat score compared with uninfected animals was 0·89 and 0·97, respectively.
Correlating bulk tank milk F. hepatica ELISA results with herd average annual milk production, while controlling for potential confounding factors resulted in an average reduction of milk yield of 0·7 kg cow−1 day−1 (≈3%) (Charlier et al. Reference Charlier, Duchateau, Claerebout, Williams and Vercruysse2007). Mezo et al. (Reference Mezo, González-Warleta, Castro-Hermida, Muiño and Ubeira2011) report a milk yield reduction in cows categorized as ‘highly positive’ based on F. hepatica (bulk-tank) milk ELISA of 1·5 kg cow−1 day−1 (≈5%) and 2 kg cow−1 day−1 (≈6%) on the herd and animal level, respectively. Finally, a randomized placebo-controlled trial administering closantel at dry-off indicated that these losses are largely recoverable as treatment resulted in a 1·1 kg increase in peak production and a longer persistence of lactation (9%) resulting in a 305-day milk production increase of 303 kg (Charlier et al. Reference Charlier, Hostens, Jacobs, Van Ranst, Duchateau and Vercruysse2012a ). There was no effect of treatment on the fat or protein concentration of the milk.
A less clear picture is obtained for the effect of F. hepatica on reproductive performance. Using a high experimental infection dose of 600 metacarcariae, Lopez-Diaz et al. (Reference Lopez-Diaz, Carro, Cadorgina, Diez-Banos and Mezo1998) found that first oestrus was delayed by 39 days in infected heifers compared with uninfected controls. This effect was linked to significantly higher oestradiol and lower progesterone concentrations in the serum of the infected animals. Loyacano et al. (Reference Loyacano, Williams, Gurie and DeRosa2002) report non-significant but higher pregnancy rates (67 vs 54%) in beef heifers under repeated flukicide control compared with untreated controls. Charlier et al. (Reference Charlier, Duchateau, Claerebout, Williams and Vercruysse2007) report that an increase in anti-F. hepatica antibody level in bulk tank milk was associated with an increase in the mean intercalving interval of 4·7 days in dairy herds. In contrast, Simsek et al. (Reference Simsek, Risvanli, Utuk, Yuksel, Saat and Koroglu2007) and Mezo et al. (Reference Mezo, González-Warleta, Castro-Hermida, Muiño and Ubeira2011) found no association between the F. hepatica serological status and repeat breeding or calving to conception interval.
In conclusion, new studies, underpinned by improved diagnostic tools, are generating insights into the production impact of fasciolosis in modern cattle systems. There is sound evidence that milk production in dairy cattle or carcass weight of slaughtered beef cattle in F. hepatica-infected herds will decrease on average by 3–5% or 0·5–0·7%, respectively. Also muscle conformation and carcass fat composition may be affected. However, studies quantifying the effect of F. hepatica on growing heifers or on reproductive performance remain limited. Current studies support deleterious effects on reproductive performance, but mainly under high infection challenge and in heifers. There is a need for randomized controlled clinical trials to assess the effects under field circumstances.
FASCIOLA HEPATICA AND CO-INFECTIONS
Fasciola hepatica has a well-defined, direct effect on the health and productivity of cattle but there is a growing awareness of the impact of infection on an animal's susceptibility to other pathogens. Infection with F. hepatica is known to be associated with a powerful anti-inflammatory immune response, dominated by an antibody response of the IgG1 isotype and a cellular response associated with the cytokines interleukin (IL)4, IL10 and Transforming Growth Factor-β (Brady et al. Reference Brady, O'Neill, Dalton and Mills1999; Flynn et al. Reference Flynn, Irwin, Olivier, Sekiya, Dalton and Mulcahy2007). Dalton et al. (Reference Dalton, Robinson, Mulcahy, O'Neill and Donnelly2013) describe three well-defined molecules, secreted by F. hepatica with immunomodulatory properties. This strongly polarized immune response appears to have consequences for the host and particularly its ability to control intracellular pathogens. For example it has been known for many years that Salmonella dublin infection in cattle is associated with the presence of F. hepatica (Aitken et al. Reference Aitken, Hughes, Jones, Hall and Collis1978; Vaessen et al. Reference Vaessen, Veling, Frankena, Graat and Klunder1998).
In mice, infection with F. hepatica increases susceptibility to the bacterial infection Bordetella pertussis and decreases the efficacy of the B. pertussis vaccine (Brady et al. Reference Brady, O'Neill, Dalton and Mills1999). This is associated with induction of a potent T helper (Th) cell type 2 response by F. hepatica and corresponding suppression of Th1 and interferon (IFN)-γ responses required to control the B. pertussis infection. These findings led researchers to investigate the interactions between fluke and other pathogens normally controlled by Th1 responses. One important pathogen affecting cattle in many countries of the world is Mycobacterium bovis, the causative agent of bovine tuberculosis (BTB). Although the question, ‘does F. hepatica infection lead to a change in the susceptibility of cattle to BTB?’ has not yet been addressed effectively, recent work suggests that at the very least, the diagnosis of BTB may be compromised in fluke-infected cattle. Flynn et al. (Reference Flynn, Irwin, Olivier, Sekiya, Dalton and Mulcahy2007) showed that cattle co-infected with the avirulent M. bovis strain, Bacille Calmette Guerin (BCG), had a significantly reduced diagnostic response to the bacterium. Diagnosis of BTB is based on the single intradermal comparative cervical tuberculin (SICCT) test, in which purified protein derivative (PPD), an antigen of M. bovis, is inoculated into the skin of a cow. If the animal is infected with M. bovis, there is a specific delayed-type hypersensitivity response, involving activation of antigen-specific T cells and secretion of IFN-γ. The magnitude of the response and associated thickening of the skin is used to diagnose infection. Suppression of this response has been demonstrated in calves co-infected with F. hepatica under experimental conditions (Flynn et al Reference Flynn, Irwin, Olivier, Sekiya, Dalton and Mulcahy2007, Reference Flynn, Mulcahy, Welsh, Cassidy, Corbett, Milligan, Andersen, Strain and McNair2009; Claridge et al. Reference Claridge, Diggle, McCann, Mulcahy, Flynn, McNair, Strain, Welsh, Baylis and Williams2012). Perhaps more significantly, these findings have been extended to cattle in the field, again enabled by improved diagnostic tests. Comparison of the spatial distribution of prevalence of F. hepatica infection and incidence of BTB in over 3000 dairy herds in England and Wales, showed almost no overlap and logistic regression analysis found presence of F. hepatica was significantly negatively associated with diagnosis of BTB (Claridge et al. Reference Claridge, Diggle, McCann, Mulcahy, Flynn, McNair, Strain, Welsh, Baylis and Williams2012). These data, together with results from experimental co-infections, suggest that fluke infection in dairy cattle suppresses the diagnostic SICCT test for BTB. This finding has important consequences for the control of BTB. It is also important to investigate the interaction between fluke and other intracellular infections that cause significant disease in cattle. Johnes disease caused by Mycobacterium avium paratuberculosis, bovine viral diarrhoea virus and Clostridium spp. are all significant pathogens of cattle whose control may be affected in animals exposed to F. hepatica. The effect of F. hepatica on vaccine efficacy is also not well understood.
Interaction between pathogens is complex and the outcome in co-infected animals may well depend on which pathogens are present. For example, whilst the immune response to B. pertussis is significantly compromised in mice co-infected with F. hepatica, the immune response to the potent Th1 inducing parasite, Toxoplasma gondii, is unaffected by the presence of a fluke infection (Miller et al. Reference Miller, Smith, Ikin, Boulter, Dalton and Donnelly2009). Much more research is required to fully understand the interaction between different pathogens and the overall outcome of such co-infections in cattle naturally exposed to a plethora of pathogens.
IMPROVING DISEASE FORECASTING
The emergence of fasciolosis in some areas, increasingly obvious production effects, and the threat of anthelmintic resistance with unfocused drug use, emphasize the need for prediction and forecasting tools at regional and farm scales. The development and death rates of the free-living stages of F. hepatica and the population dynamics of the snail intermediate hosts are strongly influenced by climate, and this underpins seasonal patterns of infection in temperate areas (Dalton, Reference Dalton1999; van Dijk et al. Reference van Dijk, Sargison, Kenyon and Skuce2010), and provides a basis for predictive models. Attempts to forecast autumnal fasciolosis risk from climatic variables such as environmental temperature and, especially, rainfall have been made for a long time and, with climate change, this field is currently attracting renewed attention. The objective of fluke forecasting is first to determine whether a farm is in a fluke risk area, second what the expected within-year fluke prevalence will be on this farm (in terms of above- or below-average risk) and third exactly when pasture contamination becomes dangerous, in order to guide treatment decisions or evasive strategies. So far, no models have been able to make successful predictions at all these levels.
Forecasting systems have been available since the 1950s (Ollerenshaw and Rowlands, Reference Ollerenshaw and Rowlands1959; Malone et al. Reference Malone, Williams, Muller, Geaghan and Loyacano1987; McIlroy et al. Reference McIlroy, Goodall, Stewart, Taylor and McCracken1990; Gaasenbeek et al. Reference Gaasenbeek, Over, Noorman and de Leeuw1992). They are all based on the use of indices of temperature and humidity in the months that most influence Fasciola epidemiology to warn farmers about above-average predicted autumnal disease risk. However, the seasonal temperature-driven developmental window for Fasciola can vary substantially between similar eco-climatic zones (Yilma and Malone, Reference Yilma and Malone1998; Bossaert et al. Reference Bossaert, Lonneux, Losson and Peeters1999). Therefore forecasting systems should be regarded as region-specific.
Recent research efforts have invariably focused on quantifying and predicting spatial risk. These efforts build further on the work of Malone and colleagues who introduced the use of geographic information systems (GIS) to study the spatial distribution of liver fluke (Malone et al. Reference Malone, Fehler, Loyacano and Zukowski1992). Recognizing the multiple causation of disease, researchers have explored risk factors other than rainfall and temperature by constructing models providing the best fit for estimates of Fasciola abundance. In the UK, McCann et al. (Reference McCann, Baylis and Williams2010b ) identified significant predictors such as soil pH and the slope of land whereas in Australia the variable that best explained spatial fluke distribution was irrigation (Durr et al. Reference Durr, Tait and Lawson2005). In Cambodia, Tum et al. (Reference Tum, Puotinen and Copeman2004) built their maps using factors including inundation risk, proximity to rivers and elevation. However, Bennema et al. (Reference Bennema, Ducheyne, Vercruysse, Claerebout, Hendrickx and Charlier2011) showed that, in addition to factors describing the parasite's environment, the inclusion of farm management factors such as the mowing of pastures and the length of the grazing season were also vital for the correct prediction of infection risk in Belgian cattle. The differing variables that correlate with Fasciola abundance in different studies show that, like the older forecasting systems, current models predicting spatial risk should be built at the national, or regional, scale and are unlikely to be applicable to other countries. McCann et al. (Reference McCann, Baylis and Williams2010a ) furthermore showed that, even using a plethora of predictors, it is as yet impossible to reliably predict risk at the farm level, with the postcode (roughly NUTS 2 equivalent) level the smallest scale achievable. It appears that GIS-based modellers face the same dilemmas that mechanistic modellers of temporal disease risk, across disciplines, have found hard to deal with (Smith, Reference Smith2011): (1) to pursue on a road of ever-increasing complexity, ultimately producing models that accurately predict risk at spatially fine scales yet not being widely applicable; or (2) to focus on the simplest model still giving valuable insights on trends, but unsuitable for the prediction of farm-level risk needed for decision support.
Knowledge of parasite ecology at different latitudes could be used to modify established models. Improved high-throughput diagnostic tests, alongside now widely available remotely sensed climatic data, make it ever more feasible to develop statistical models of spatio-temporal risk. However, changes to the underlying epidemiological processes that are implicitly, not explicitly, captured in such models would limit ability to extrapolate conclusions and forecasts between regions or into future climates. Changes in seasonality and the relative importance of rainfall and temperature in driving parasite abundance and availability are core to this problem. Fortunately, improved diagnostics also have the potential to better validate mechanistic models of the ecology of the free-living stages, including the snail host, which could explain regional differences and be robust to future changes. In general terms, since it is likely that rainfall is an important determinant of annual risk whereas especially temperature determines when pasture becomes dangerous (Gettinby et al. Reference Gettinby, Hope-Cawdery and Grainger1974), the effects of both should be included in such models. A first challenge will be to quantify the effects of temperature on, notably, survival of eggs at pasture, the size of snail populations, as well as individual snails, and metacercaria. A second, major, challenge, shared with models of many other parasites, is to identify parameters which link rainfall to the microclimate experienced by the parasite and snail as well as relate them to their developmental success. A third challenge will be to quantify the spatial distribution of snail habitats on farms. In the longer term, after an assessment of the variance in crucial rates at various spatial levels, these models should also incorporate parasite adaptation. As the life cycle of the parasite is complex, it will be essential to use sensitivity analyses to identify key parameters and strip prediction models down to the key drivers (Smith, Reference Smith2011). This would enable models to be run with limited farm-level input data, on mobile platforms, and hence enhance their value in decision support.
CONCLUSIONS AND PERSPECTIVES
Our review highlights that fasciolosis should be considered an important production disease of cattle in Europe. In the countries that report results, prevalences remain high, despite ongoing control efforts and renewed attention to the disease. In several areas significant scientific progress has been made over the last decade such as the development of diagnostics, understanding and prediction of spatial distribution of disease occurrence and effects of fluke on susceptibility to co-infection and diagnosis of other infections. In other areas such as the development of new flukicidal compounds little progress has been made (Fairweather and Boray, Reference Fairweather and Boray1999) and therefore has not been included in this review. In Fig. 1, we identify the key areas for future research. As stated above, major progress could come from new anthelmintic medicines. The major bottlenecks for current flukicides are (1) their limited activity against immature stages; (2) the prohibition of their use in animals producing milk for human consumption; and (3) the development of anthelmintic resistance. The development of new medicines that overcome some or all of these issues would be a major advancement. The low discovery and development rate of new medicines may to a large extent be explained by low economic incentives to the pharmaceutical industry. However, with the increasing concerns about fasciolosis the balance is tilting the other way and investment in new drug development is needed. In addition, the European Commission and the pharmaceutical industry are making considerable efforts to promote research into vaccine development (Golden et al. Reference Golden, Flynn, Read, Sekiya, Donnelly, Stack, Dalton and Mulcahy2010) and establish maximum residue limits for all flukicidal drugs (European Commission Implementing Decision C(2012)8604), potentially allowing their use in dairy cattle.
Fig. 1. The key areas for future research to promote economic management of bovine fasciolosis.
Some topics are receiving renewed attention such as the biology of intermediate host and free-living stages. Better understanding of the climatic and environmental conditions on their propagation/survival is key to improved temporal and spatial prediction of disease occurrence.
Development of diagnostics, which has contributed much already to advances in understanding and application, should further focus on the relationship between morbidity and productivity in different epidemiological settings. The production impacts need to be further defined in growing cattle. Better understanding in this area is also expected from randomized clinical field trials that include assessment of fertility parameters. The evidence that F. hepatica modulates the immune response and affects the susceptibility and diagnosis of co-infections is an area that requires further elucidation, particularly where those diseases are part of a national control programme. A better understanding of the overall impact of fluke infection will allow close collaboration with agricultural economists and will place fasciolosis control next to other key farm management and animal health decisions in the whole-farm economic context. Ultimately, this would mean that we would stop controlling the parasite and start managing it.
FINANCIAL SUPPORT
This work has received funding from the European Union Seventh Framework Programme [FP7/2007-2013] under Grant agreement No. 288975CP-TP-KBBE.2011.1.3-04; GLOWORM.
CONFLICT OF INTEREST
The SVANOVIR®F.hepatica-Ab ELISA is commercialized by Boehringer Ingelheim Svanova under a licence agreement with Ghent University.
