Angiostrongylus vasorum infection in dogs: continuing spread and developments in diagnosis and treatment
Disease caused by Angiostrongylus vasorum is increasingly diagnosed in dogs, as the geographic range of the parasite increases along with awareness among clinicians. Diagnosis, treatment and prevention are not always straightforward, although recent developments offer hope for improved options in future. Understanding of the epidemiology and pathogenesis remains poor. This paper provides an overview of the current state of knowledge of this parasitic disease, focussing on the most recent developments and advances.
The metastrongyloid nematode Angiostrongylus vasorum is a parasite of the heart and pulmonary circulation of dogs and foxes. Infection can cause a wide range of disease outcomes, which are most often characterised by respiratory dysfunction, but can also manifest as bleeding, neurological, cardiovascular or gastrointestinal disorders, with or without respiratory involvement. Although well known for some time, in recent years the parasite has been increasingly reported in dogs and foxes, in association with geographical range expansion and possibly increasing transmission in endemic areas. The purpose of this review is not to provide comprehensive information on the parasite and associated disease, since this has been recently reviewed elsewhere (Koch and Willesen 2009, Helm and others 2010), but to provide an update on the areas where our knowledge and understanding haveprogressed in the last few years. These areas primarily concern changes in distribution, and developments in diagnosis and treatment. The main aspects of life cycle, pathogenesis and epidemiology are summarised, but advances in understanding of these aspects have been slower to materialise.
Life cycle and pathogenesis
The basic life cycle of A. vasorum is well described (Bolt and others 1994), although doubts persist over some details (Morgan and others 2005). Adults of both sexes are found in the right side of the heart and the pulmonary arterial circulation, and mate to produce eggs, which hatch quickly. First-stage larvae (L1) are carried to the pulmonary capillary bed, penetrate into the alveoli and are carried to the pharynx, swallowed, and pass out in the faeces. Further development to the infective third larval stage (L3) takes place in a gastropod mollusc (slug or snail) intermediate host. On being ingested by the definitive host, L3 migrate via the blood and/or lymph to the heart and pulmonary arteries, where they mature.
Early experimental work established that the potential intermediate host range is broad, with most species of gastropod tested proving suitable for development of L3. In nature, the range of species found to be infected in endemic areas is also very broad (Ferdushy and others 2009), although the relative importance of different species to transmission is likely to vary. Infection occurs when gastropods feed on infected faeces. Different gastropod species vary in their propensity to feed on faeces, such that some are more likely to become infected than others.
The way in which most definitive hosts become infected is unclear. It is assumed that ingestion of gastropods containing L3 is the primary route of infection. This could be by predation, but also by inadvertent ingestion through, e.g., scavenging on carrion, or grass-chewing. Activity patterns of slugs and snails in relation to dogs’ behaviour and habitat use are therefore likely to underlie individual risk factors for infection, as well as epidemiology at the population level. Experimentally, L3 have been stimulated to leave snails, and survive for several days thereafter (-Barcante and others 2003), so it is possible that dogs could be infected by ingestion of free L3 from the environment. However, this work used an unnatural host, the tropical water snail Biomphalaria glabrata, kept in unnatural conditions, and larvae emerged into water. The relevance of this route to natural transmission therefore remains uncertain. The potential role of paratenic hosts such as frogs was signalled some time ago (Bolt and others 1993), but their actual role in transmission to dogs is not known.
It is worth noting that much experimental work on the intermediate and definitive hosts has been conducted in South America using local parasite isolates, which are genetically distinct from A. vasorum in the northern hemisphere (Jefferies and others 2009a). Although not yet demonstrated, biological differences could exist and limit the extent to which these data can be extrapolated to parasite populations in Europe and North America.
The main definitive hosts of A. vasorum are dogs and foxes. In Europe, the red fox Vulpes vulpes is assumed to be the only significant wild host (Saeed and others 2006), while the pampas fox Pseudalopex gymnoceros, hoary zorro (Pseudalopex vetulus) and crab-eating fox (Dusicyon thous) are infected in South America (Bolt and others 1994). Recent work has documented infection in coyotes in North America (Bourque and others 2005, Bridger and others 2009). Infections in red pandas in zoological collections (Patterson-Kane and others 2009) and in a wild stoat in England (Simpson 2010) also suggest that the definitive host range might be broader than that first appreciated. A. vasorum has been described in badgers (Meles meles) (Torres and others 2001). However, badgers are also commonly infected with Aelurostrongylus species, so there is potential for misidentification in coprological surveys. It is likely that as yet undescribed species of lungworm exist in various wildlife species, and careful morphological and molecular confirmation of A. vasorum is therefore important before considering new wildlife species to be hosts and potential reservoirs. Cats have been experimentally infected with South American A. vasorum, although few worms established and no larvae were produced (Dias and others 2008). Natural infections of cats with A. vasorum have not been reported. Cats host the lungworm Aelurostrongylus abstrusus (Traversa and Guglielmini 2008), whose first-stage larvae closely resemble those of Angiostrongylus, and this could lead to coprological mis-diagnosis.
Various pathological changes in the lungs have been described in dogs and foxes naturally or experimentally infected with A. vasorum (Bourque and others 2008, Willesen and others 2008a, Schnyder and others 2009a,b). In chronic cases, these were associated with infiltration of inflammatory cells, disruption of alveolar architecture, thrombus formation, and, later, fibrosis (Bourque and others 2008), which occurred especially around larval nematodes. Similar changes were seen in naturally infected red pandas (Patterson-Kane and others 2009). It is therefore possible that the host response to L1 penetration of the alveoli underlies much of the pulmonary pathogenesis. However, in recent experimental infections of dogs, lung pathology was demonstrated during the prepatent period (Schnyder and others 2009a), so mature infections might not be necessary for respiratory disease to occur.
Mechanisms of the bleeding disorders associated with A. vasorum remain unknown, with most studies suggesting a consumptive coagulopathy, possibly related to chronic, low-grade disseminated intravascular coagulation (Cury and others 2002, Adamantos 2009, Denk and others 2009, Schmitz and Moritz 2009, Helm and others 2010). Further work is needed in this area before management of bleeding disorders in A. vasorum infection can be based on sound evidence. Individual host genetic factors could be important, given the wide variety of clinical presentations in practice (Koch and Willesen 2009). In spite of widespread larval migration potentially causing pathology in a range of organs, most clinical diseases outside the cardio-pulmonary system are likely to be related to bleeding disorders, e.g. neurological signs secondary to bleeding into the central nervous system (Denk and others 2009). Bleeding can also occur into other organs or body cavities (Sasanelli and others 2008, Willesen and others 2008b). In spite of the traditional, archaic name French heartworm, cases of A. vasorum rarely present with cardiac disease (Koch and Willesen 2009).
Since being described in the south of France in the mid-19th century, A. vasorum has been reported as a parasite of foxes, and a cause of disease in dogs, in an increasing number of countries in Europe and beyond. This distribution is patchy, with apparently stable endemic foci surrounded by areas of low prevalence (Morgan and others 2005, Jefferies and others 2010). In recent years, however, there is evidence for spread beyond these traditional endemic areas. In the UK, for instance, the parasite was reported in south Wales and south-west England in the 1980s, in south-east England in the 1990s, and then seems to have spread north, with new cases in non-travelled dogs in the north of England and Scotland in the past two years (Helm and others 2009, Yamakawa and others 2009). This predominance in southern Britain, with lower prevalence further north, is also evident in foxes (Morgan and others 2008).
Range expansion is also apparent in other countries. In Denmark, the initial focus of infection in and around Copenhagen has enlarged to include other parts of the country, as well as Sweden (Åblad and others 2003), and has similarly been accompanied by expansion within the fox population (Saeed and others 2006). Following initial reports in Germany (Barutzki and Schaper 2003), laboratory data indicate that the parasite is now widespread (Barutzki and Schaper 2009, Taubert and others 2009), and first reports of infection in non-travelled dogs have been published in the last few years from the Netherlands (van Doorn and others 2008) and Switzerland (Staebler and others 2005). In Newfoundland, Canada, long-standing infection in foxes has extended to dogs (Bourque and others 2008) and coyotes (Bridger and others 2009). In Italy, cases in dogs have only recently been reported despite historical evidence of infection in foxes (Sasanelli and others 2008, Traversa and others 2008). While heightened awareness among clinicians and researchers has probably contributed to these increasing reports, the rate of increase and the parallel expansion of infection in fox populations suggest that this is truly an emerging disease. It is difficult to base such a judgement on canine clinical data alone, since there is no mechanism for central recording of cases that would enable objective between-year comparison in the extent of disease.
The factors underlying expansion in the range of A. vasorum, and increasing disease in dogs, are poorly understood. Effects of climate change on the abundance, seasonal breeding patterns and activity of the intermediate hosts, changes in fox abundance and distribution, including urbanisation, and parasite translocation through dog movement, might all be involved. Simple climatic modelling suggests that the potential geographical range is much greater than the area currently occupied by the parasite, even without invoking climate change (Morgan and others 2009). Acceleration of range expansion through dog movement might therefore explain recent spread. Local movement of foxes and carriage of intermediate hosts (Routh 2009) could also contribute. Recent work shows that while South American isolates of A. vasorum are likely to have evolved independently in ancestral canid hosts (Jefferies and others 2009a), the North American colony in Newfoundland represents a subset of the genetic diversity found in Europe and was probably introduced in historical times with dogs, foxes or intermediate hosts (Jefferies and others 2010). Certainly, there are several regions of the world that are probably suitable for A. vasorum establishment that have not yet been colonised, e.g. coastal North America, Japan, and parts of South Africa, Australia and New Zealand (Morgan and others 2009). Movement of untreated dogs to those areas from endemic countries could therefore present a serious threat to canine health and welfare.
Within endemic areas, there is anecdotal evidence for an increase in cases in dogs, suggesting improved conditions for transmission. In parts of Europe, increasing fox populations, especially in towns, have been speculatively linked to increasing angiostrongylosis in dogs. However, evidence from the UK suggests otherwise, since urban foxes have been present in many towns for decades, while their numbers and population density decreased rather than increased during the 1990s as a result of sarcoptic mange (Soulsbury and others 2007). Suitable intermediate host species, meanwhile, have long been ubiquitous in the UK, as in almost all of Europe. Local abundance, and seasonal breeding and activity patterns of gastropods are likely to be affected by climate change, which could therefore improve conditions for parasite development as well as increase exposure of dogs to infected intermediate hosts. Although responses will be species-specific, milder winter temperatures and increased precipitation are likely to generally favour slugs, and could drive expansion of some species (Willis and others 2006). Given ongoing climate change, well-established fox populations, and increasing movement of dogs between and within countries, it seems likely that exposure of dogs to infection will increase rather than decrease in the future, in endemic areas as well as potentially in those suitable but not yet colonised.
In endemic areas, where primary care clinicians are familiar with angiostrongylosis, most diagnoses are made on clinical grounds. The symptomatology of angiostrongylosis has been extensively reviewed (Koch and Willesen 2009, Helm and others 2010), and is not repeated here. Classically, coughing in a young dog is highly suspicious, but the wide age range and diverse pathologies mean that the disease should be considered as a differential diagnosis in many clinical presentations. Bleeding disorders and other less usual signs can occur with or without respiratory disease. Although published cases are likely to be biased towards more interesting and perhaps less common clinical outcomes, there is abundant evidence for diverse presentations at primary care level (Koch and Willesen 2009, Morgan and others, in press). There are limited data on risk factors for infection, but these appear to include age (younger dogs more likely to be infected) (Chapman and others 2004, Barutzki and Schaper 2009, Morgan and others, in press), recent worming history (Morgan and others, in press) and breed (Staffordshire bull terriers and cavalier King Charles spaniels marginally over-represented) (Chapman and others 2004). Diagnoses are more common in winter and spring (Conboy 2004, Taubert and others 2009, Morgan and others, in press). This could result from the increased risk of infection in late summer, when gastropods are most abundant and larval infections most mature, with disease developing over subsequent months. The seasonal pattern is weak, however, and disease can be observed at any time of year.
Radiography often shows alveolar, interstitial or mixed patterns, especially in the peripheral lung fields (Boag and others 2004). Blood biochemistry is inconsistent, with slightly decreased serum fructosamine and moderate hyperglobulinaemia observed in some cases (Willesen and others 2009). Hypercalcaemia with polyuria and polydipsia (perhaps as a result of thoracic granulomata) has been reported, and angiostrongylosis should be borne in mind as a treatable differential diagnosis for hypercalcaemia (Boag and others 2005). Peripheral eosinophilia is occasionally observed (Willesen and others 2009), along with anaemia and thrombocytopenia (Cury and others 2002). Changes in coagulation parameters are inconsistent (Schelling and others 1986, Cury and others 2002, Schnyder and others 2009b, Willesen and others 2009), perhaps reflecting the complexity of underlying host mechanisms.
Radiography and laboratory investigation can therefore be helpful, but are never pathognomonic. Definitive parasitological diagnosis relies on observation of first-stage larvae (L1) in the respiratory tract or faeces. L1 are sometimes observed on airway cytology. Broncho-alveolar lavage is therefore a diagnostic option (Barcante and others 2008), but must be weighed against the risks of the procedure in dyspnoeic dogs. Larvae and larval tracts have been reported in the eye (Manning 2007). The method of choice for recovery of L1 from faeces is the Baermann test (Koch and Willesen 2009), in which larvae migrate out of the faeces and sediment to the bottom of a receptacle, where they are concentrated and examined microscopically. The Baermann test is usually read after 8 hours or more, but larvae begin to emerge much sooner, and the test may be positive after 30 minutes in heavily infected dogs (Morgan, unpublished data). Larvae can also be detected on direct faecal smears, with a sensitivity of up to 67% compared with the Baermann test (Humm and Adamantos 2010). Faecal flotation using saturated sodium chloride solution will not reliably recover L1, which are too dense to float unless much higher specific gravity solutions are used, in which case osmotic deformation of the larvae renders specific identification difficult (Traversa and Guglielmini 2008). Specific identification of L1 is worthwhile in order to distinguish them from other lungworms such as Crenosoma vulpis, as well as free-living nematodes contaminating the sample. A. vasorum larvae are most easily identified by their characteristic tail morphology (McGarry and Morgan 2009).
A major limitation of the Baermann test is that L1 excretion appears to be intermittent (Oliveira and others 2006), and infected dogs might be negative on faecal examination. A single Baermann test is likely to detect at most 50% of infected dogs, and although sensitivity can be increased by serial examination, e.g. of samples collected three days in succession (Koch and Willesen 2009), this is not always practical. L1 survive well in refrigerated faeces for a week or so, and serial samples can be stored and pooled for examination on the third day. Given the limitations of faecal diagnosis, it is justifiable to treat dogs with a high clinical suspicion of angiostrongylosis (based on compatible history, presenting signs, and radiographic findings) if Baermann test results are negative.
Improved diagnostic tests are likely to emerge. Parasite proteins or DNA can be detected in the blood using sandwich ELISA (Verzberger-Epshtein and others 2008, Schnyder and others 2009b), or polymerase chain reaction (PCR) (Jefferies and others 2009b), respectively. The source of the antigen or DNA detected by these tests is not known, and, if from eggs or larvae, intermittent egg shedding might be expected to negatively affect test sensitivity. Serological ELISA tests to detect host antibodies to the parasite are also in development (Jefferies, unpublished data), but their usefulness in a clinical context will depend on the duration of antibody persistence and levels of exposure in the general population. Evaluation of these novel technologies for detection of A. vasorum infections is at a preliminary stage, and data on their application in a clinical context are not yet available.
With current technology, the best approach to diagnosis in practice is to use information from clinical signs and supportive data from haematology, biochemistry, radiography, parasitology and local epidemiological knowledge. Where a quick result is desirable, direct faecal smear and in-house Baermann tests, read after an hour, can be used while sending samples to an external laboratory for confirmation.
Treatment and prevention
At the time of writing, two drugs are licensed for the treatment of A. vasorum infection, while a third is unlicensed but has been in common use for some time (Table 1). Milbemycin oxime, when given four times one week apart soon after patency, reduced adult worm burdens by around 85% in experimentally infected dogs (Conboy and others 2004). A similar effect was achieved by two doses during the prepatent period (one and two months after infection), while a single dose a month after infection reduced adult worm burden by around 55% compared with controls (Conboy and others 2004). Moxidectin given either 4 or 32 days after infection prevented establishment of adult parasites (Schnyder and others 2009a). Moreover, dogs treated 4 days after infection did not develop any detectable pathology, and those treated 32 days after infection had substantially reduced pathology compared with untreated controls examined around 2 months after infection (Schnyder and others 2009a).
Table 1. Anthelmintic drugs currently licensed for, or in widespread use in, the treatment of Angiostrongylus vasorum infection in dogs
|Fenbendazole (off label use)||Panacur, Intervet-Schering Plough Animal Health||25-50 mg/kg orally once daily for 7-21 days|
|Milbemycin oxime||Milbemax, Novartis Animal Health||0·5 mg/kg orally once weekly for 4 weeks|
|Moxidectin||Advocate, Bayer Animal Health||Minimum 2·5 mg/kg topically (0·1 ml/kg of 2·5% spot-on), single dose|
Information on treatment of natural infections, beyond short case series, is limited. In 16 dogs in Canada that had respiratory disease and patent A. vasorum infection, and were treated with milbemycin oxime (protocol in Table 1), clinical signs resolved in 15 and larval shedding ceased in 14 (Conboy 2004). In a prospective trial in Denmark, 23 out of 27 dogs with patent infection that were treated with moxidectin were no longer shedding larvae 42 days later, compared with 21 out of 23 dogs treated with fenbendazole (Willesen and others 2007). In those dogs still positive after 42 days, a single additional treatment with moxidectin or milbemycin oxime eliminated larvae, while treated dogs generally showed radiological improvement. Therefore, it appears that licensed drug protocols substantially reduce worm burdens, and attenuate associated pathology. However, continued monitoring after treatment is advisable, with repeated treatment as necessary.
Supportive treatment might be appropriate on a symptomatic basis (Helm and others 2010), although no good evidence is available on its efficacy. Corticosteroids are sometimes used to prevent adverse reactions to killed worm antigen, and to reduce fibrosis in recovering lungs, though the importance of these effects has not been established. Antibiotics are not routinely indicated. Cage rest is recommended for dyspnoeic cases, and administration of oxygen-enriched air might be of benefit, along with diuretics and bronchodilators (Koch and Willesen 2009). In the absence of better understanding of the mechanism of the bleeding disorders, it is impossible to make specific recommendations, other than whole blood transfusion or component therapy for cases with major haemorrhage. Although mortality rates appear to be high among the severe cases reported in the literature, at primary care level the prognosis is excellent for most respiratory cases (Willesen and others 2007), and where death occurs it is usually in dogs with severe coagulopathy (Koch and Willesen 2009). Logically, treatment is more likely to be successful with early anthelmintic therapy, which, given the poor sensitivity of larval detection, should not necessarily be delayed until a definitive diagnosis is reached.
At the time of writing, there were no drugs licensed for the prevention of A. vasorum infection. Treatment in the prepatent period with milbemycin oxime (Conboy and others 2004) or moxidectin (Schnyder and others 2009a) decreases or prevents establishment of adult parasites. Some pathology seems to develop during the prepatent period (Schnyder and others 2009a), so re-infection after treatment could potentially lead to disease even in regularly treated dogs.
It is not known to what extent dogs are protected from re-infection by the persistence of macrocyclic lactone anthelmintics, nor how severity of disease relates to level or stage of infection. This makes it very difficult to make rational recommendations for prevention of infection, or clinical disease, in practice. Regular treatment with an effective drug would logically be expected to reduce the risks, and more regular treatment should reduce the risks more. Dogs in an endemic area treated every three months with milbemycin oxime were about half as likely to test positive for A. vasorum as those given fenbendazole or untreated (Morgan and others, in press). The level of parasitological efficacy needed to confer clinical protection is not known. Certainly, clinicians should be vigilant for A. vasorum infection even in dogs being treated regularly with anthelmintics. Seasonal patterns of infection (see above) could in theory help to target treatment at high risk times of year, but from the current evidence it is not possible to identify a distinct period of high risk. Recommendations for preventive treatment should also take account of the local epidemiological situation. Intensive anthelmintic prophylaxis targeted specifically at A. vasorum might not be warranted where the parasite is rare or absent. Excessive treatment is expected to favour the development of anthelmintic resistance, although this has not been reported in A. vasorum.
Avoiding ingestion of intermediate hosts is likely to decrease risk of infection, but it is difficult to see how this might be achieved. Application of molluscicide to dogs’ environments is unlikely to lead to general reductions in gastropod population size and might be counterproductive, not only in terms of molluscicide toxicity to dogs, but also by increasing access of dogs to killed slugs and snails. Given that L3 released from intermediate hosts can survive for some days in the environment (Barcante and others 2003), and that living L3 are routinely extracted from autolysed slug tissue in the laboratory (unpublished data), such killed gastropods may harbour viable parasite larvae. Other non-toxic methods of slug and snail control, e.g. baited traps, might also increase exposure while having limited or no impact on total slug populations. Decreasing L1 availability to slugs through biological control with predatory fungi is possible (Braga and others 2009) but of uncertain practical value in terms of control. Some management of the environment and exercise regimen to decrease contact with slugs and snails might be appropriate. However, the ubiquitous nature of the intermediate hosts means that, in endemic areas, regular anthelmintic treatment is probably the main form of defence.
A. vasorum is an emerging parasite of dogs, as it spreads from known endemic foci. Clinician awareness is an important defence against this disease, which can be severe and difficult to diagnose definitively. Although the pathogenesis is not fully understood, and current tests for parasitological diagnosis suffer from poor sensitivity, signs associated with infection are well described and should alert clinicians to the possibility of infection. Climatic models suggest that there is substantial scope for further geographical spread, so practitioners in areas not currently considered endemic should be vigilant. There are effective drugs for the treatment of A. vasorum, and early anthelmintic therapy of uncomplicated respiratory cases of mild or moderate severity has excellent prognosis for full recovery. More severe or unusual presentations, including bleeding disorders, can present a greater challenge. Regular anthelmintic treatment is likely to reduce, but might not eliminate, the risk of disease. While increasing numbers of dogs will be at risk of infection as the geographical range of the parasite expands, better epidemiological understanding and increasing options for diagnosis and treatment have the potential to improve prospects for disease prevention and management in the longer term.
Conflict of interest
None of the authors of this article has a financial or personal relationship with other people or organisations that could inappropriately influence or bias the content of the paper.