“Gone in the back legs”

Neural angiostrongyliasis and neosporosis are the two most common infectious causes of spinal cord disease in young dogs. The former is caused by migration of Angiostrongylus cantonensis (rat lungworm) larvae following ingestion of mollusc intermediate hosts, while the latter is caused by the apicomplexan protozoan Neospora caninum, acquired transplacentally, during parturition or in the neonatal period. This article gives the reader the perspective of a veterinarian confronted with the diagnosis and treatment of these two potentially life-threatening infections, taking into account differences in epidemiology and pathogenesis, and listing diagnostic tests available and affordable for most owners. The broader implications of these infections for other species, including people and wildlife, are discussed.

Most readers of Microbiology Australia are scientists. In this article, we want you to leave the comfort zone of the laboratory and embrace the world of companion animal medicine. An old proverb says: to understand a man, you’ve got to walk a mile in his shoes, whether they fit or not. So metaphorically, we ask you to “wear” the shoes of a suburban veterinarian asked to look at a young dog who has “gone in the back legs”, as many owners would put it. First we need to give you the rules of engagement, because although a profession, veterinary science is also a small business where owners expect an accurate diagnosis and a successful cost-effective solution. Good outcomes are expected even if the patient is seen at nights or at weekends, when laboratories do not offer a comprehensive service.

The dog in question is a 12-week-old Golden Retriever called “Honey” normally domiciled in Bellevue Hill (Fig. 1). She has a history of progressive hind limb dysfunction over several days. The onset was sudden, without evidence of trauma. Neurologic testing shows poor proprioceptive ability in the hind limbs, weakness that seems to be lower motor neuron (LMN) in type with poor tail carriage, urinary and faecal incontinence and marked hyperesthesia over the tail base and adjacent spine. This unusual combination of signs is most suggestive of a disease affecting the spinal cord (especially the part subserving the back end of the dog), or the corresponding nerves and/or nerve roots innervating the legs, tail and rear-quarters. Hyperaesthesia suggests meningeal inflammation. Not many diseases produce this combination of signs and common diseases such as tick paralysis (due to the toxin of Ixodes holocyclus) and traumatic vertebral injury do not fit the clinical picture (trust us, we are vets!). Congenital vertebral anomalies are rare in this breed and vertebral or spinal cord neoplasia is extremely rare in young dogs. So, the two most likely possibilities are neural angiostronglyosis (NA)^1–5 and neosporosis^6–11, two parasitic diseases which seem disproportionately common in eastern Australia compared to the rest of the world.

Figure 1. A 3-month-old female golden retriever presented for hind limb weakness and urinary/faecal incontinence. Eosinophilic pleocytosis and anti-A. cantonensis antibodies in CSF demonstrated that this dog had neural angiostronglyosis. The dog recovered completely following corticosteroid treatment and symptomatic and supportive care.

Now if a young child was presented with these signs, the investigation would likely involve a neurological consultation, perhaps an infectious diseases consultation, magnetic resonance (MR) imaging of the spinal cord and brain, electrodiagnostic tests (needle electromyography and nerve conduction studies), collection and analysis of cerebrospinal fluid (CSF), blood cultures and various serological tests^12–14. Such a “work up” is possible for canine patients but many owners are unwilling to pay thousands of dollars to cover the cost of investigations. Unfortunately, Medicare doesn’t cover dogs and most people don’t take out pet health insurance. Welcome to the world of your local vet! To better understand the most likely diagnostic possibilities, to help decide which tests are most efficient, let’s consider the epidemiology and pathogenesis of these infections.

Neural angiostronglyosis is a disease caused by migration of larvae of the rat lungworm Angiostrongylus cantonensis^1–5. It generally occurs after a dog, typically (but not invariably) a young dog, ingests slugs or snails (Fig. 2). Anecdotal evidence from Sydney and Brisbane indicates infection of non-native rats (Norwegian and black rats) with this parasite is widespread^15,16. Accordingly, mollusc intermediate hosts have a very high probability of harbouring infective larvae. Thus when taking a history, enquiring whether rats, snails and slugs are present in the dog’s environment is pertinent, as often owners will have seen the patient ingest molluscs. The most characteristic feature of these cases is the exaggerated pain response to palpation and manipulation of the tail base, hind limbs and spine⁵. The worst cases experience excruciating pain. Signs tend to ascend, so some dogs present for nuchal rigidity referable to neck pain from parasitic meningitis, while the hind limb weakness changes neurologically from a LMN type to a mixed upper motor neuron (UMN) and lower motor neuron paresis. Later, a variety of other neurologic abnormalities, including cranial nerve palsies, may develop⁵. The prognosis with treatment is usually favourable, although some cases are left with residual neural deficits which occasionally result in euthanasia^1–3,5.

Figure 2. Life cycle of Angiostrongylus cantonensis^1,2,5. The natural life cycle alternates between rats and molluscs (slugs or snails). The life cycle ends in dogs (blue line) because the ingested larvae very rarely mature into adults and no offspring are produced – it is a “dead end” for the parasite and sometimes for the host as well!

Neosporosis is a protozoan disease caused by Neospora caninum^6–11, an apicomplexan closely related to Toxoplasma gondii (the cause of toxoplasmosis)¹¹ and Sarcocystis neurona (the cause of equine protozoan myelitis, a disease of horses in North America)^17,18. Neospora caninum has a very similar life cycle to T. gondii; however, the dog, coyote, wolf or dingo (all genetically the same species of canid) are definitive hosts for N. caninum, whereas the cat is the definitive host for T. gondii¹¹ and the opossum for S. neurona¹⁷. Neosporosis is mostly seen in young dogs (typically less than 6 months) infected via their dams, most likely in late pregnancy, during parturition or in the early neonatal period during lactation^6,7; the exact mechanism is not established. After a variable latent period, pups develop signs of an ascending radiculomyelopathy, initially manifest as LMN-type hind limb weakness. If this is not diagnosed in a timely manner, the spinal cord and nerve roots subserving the hind limbs are damaged irreparably, resulting in rigidity and contracture of the pelvic limbs (Fig. 3). Spinal hyperaesthesia is usually not a conspicuous feature, although some cases have myalgia and lameness due to concurrent protozoan myositis. Pups with neosporosis can be completely cured if therapy is started early; delayed therapy results in permanent and severe neurologic deficits.

Figure 3. A Boxer pup with neosporosis. This patient was presented when there was already some contracture of the left hind limb (arrow). Treatment at this stage can arrest disease progression but the permanent loss of motor neurones means complete recovery of hind limb function is not achievable.

What clinical features are most helpful in distinguishing neural angiostrongyliasis and neosporosis? Neosporosis appears to be more common in Boxers and Labradors than other breeds, early in the disease LMN features predominate (viz. reduced muscle tone and myotatic reflexes in the hind limbs) and hyperaesthesia is not prominent. In contrast, hyperaesthesia is the conspicuous feature of NA, and the neurologic picture changes from LMN to UMN as larvae migrate “up” the spinal cord towards the brain. In both diseases, multiple littermates can be affected. In NA this is because littermates may share an environment where rats and molluscs are abundant, whereas in neosporosis siblings are affected as infection is vertically transmitted^6,7. Both diseases are more common in large breeds^5,6. Time of presentation can be helpful for NA, as this condition tends to be more common in autumn⁵, although cases can occur in any month. Finally, geography can be an important consideration because to date NA has not been reported in Melbourne, Adelaide or Perth, although it occurs from Sydney all the way north along the coast of eastern Australia^1–5,15,16.

Two tests are most helpful in differentiating between these infections. NA is accompanied by eosinophilic meningoencephalitis, so finding eosinophilic pleocytosis in CSF is strongly suggestive of this aetiology^1–5. The diagnosis can be confirmed by demonstrating antibodies against A. cantonensis in CSF using an ELISA^4,5 developed by Rogan Lee at Westmead Hospital. In dogs, CSF is generally collected from the cisterna magna under general anaesthesia. Myelography, computed tomography myelography or MR imaging can be conducted at the time of CSF collection. To date, MR studies have not been helpful in the specific diagnosis of NA in canine patients (in contrast to human cases where the presence of larval migration “trails” can be sometimes detected⁹). However, imaging studies exclude most other diagnostic possibilities. Neosporosis is normally diagnosed using an immunofluorescent antibody test (IFAT); by the time dogs are presented, they have elevated serum antibody titres against this protozoan. The IFAT is only available in laboratories in Perth (VetPath) and Launceston (Mt. Pleasant Laboratory) and assays are typically only run once or twice a week, so there is invariably some delay in obtaining a result. In neosporosis, CSF cytology shows a mixed pleocytosis, often with a prominent neutrophilic component and very rarely zoites have been recorded in CSF after cytocentrifugation^6,7.

The conundrum for the vet is that treatments of these two diseases are quite different. Neosporosis is treated with antimicrobials effective against apicomplexan protozoans, traditionally trimephoprim-sulphadiazine, pyremethamine and clindamycin^6,7. Usually two of the three drugs are selected. The combination of sulphadiazine and pyremethamine has a theoretical advantage due to sequential attacks on the parasite’s folic acid pathway. In the future, drugs such as toltrazuril and ponazuril may also prove to be effective¹⁸, the latter having become the drug of choice for treating equine protozoan myelitis.

In contrast, NA is traditionally treated with corticosteroids to dampen the eosinophilic inflammatory response until the larvae die (because the dog is the “wrong” host)⁵. Perhaps we should be braver and actually kill the larva using an anthelmintic after starting aggressive corticosteroid therapy, choosing a drug regimen that kills larvae slowly using albendazole or fenbendazole, to avoid a sudden release of metazoan antigens. Such an approach has been used in human patients in Asia¹⁹ and more recently in two patients at Westmead Children’s Hospital²⁰. Dogs with neosporosis treated with corticosteroids deteriorate dramatically. Accordingly, many veterinarians treat suspect cases (such as “Honey”) with anti-protozoan drugs while awaiting serology for neosporosis, or “bite the bullet” and collect CSF to look for eosinophilic pleocytosis, if owners are agreeable. Where owners do not permit or cannot afford the full investigation, a presumptive diagnosis of NA is made after excluding neosporosis by serology or a treatment trial; corticosteroids are then administered.

It is tantalising to speculate why clinical neosporosis is apparently more prevalent in Australia than in other countries. One possibility is that it is more common to feed dogs uncooked kangaroo and beef meat in Australia (sold as “pet mince” at supermarkets) compared to other countries. Such fresh meat might contain viable cysts of both N. caninum (and T. gondii) sufficient to establish subclinical disease in the bitch, which would be transmitted subsequently to pups.

It is also fascinating that the only reports of NA in dogs have been from Australia, even though the parasite exists throughout Asia, the Pacific islands and in some southern states of the USA. This condition is behaving as an emerging infectious disease in Australia, having spread from south eastern Queensland where it was first reported by Mason in the early 1970s^1,2, reaching Sydney in 1991³ and becoming progressively more common there¹⁶. Nineteen new canine cases have been diagnosed in a single referral centre in Sydney over the past 4 years (E. Thrift, A. Lam and G. Child, personal communication). It is of great interest that NA has been reported in a wide range of species other than the dog - horses²¹, multiple animals in zoological collections (especially primates)^22,23, macrobats²⁴, possums²⁵ and tawny frogmouths^25,26. From a population ecology standpoint, impact on tawny frog mouths is greater than for any other species^25,26.

From a human perspective NA is zoonotic. It occurs in patients who (accidently or purposely!) eat slugs, snails or planarians. The most celebrated Australian case was a silly chap who ate a slug on a dare at a bucks night¹³. Far more serious is the threat posed to young infants, who have the propensity to put anything into their mouths, including molluscs. An immature immune system and small spinal cord means the neurologic impact on children to be far worse than in adult patients, potentially resulting in death or permanent sequelae. For this reason, we should do much more - as scientists, doctors, veterinarians and parents – to (i) alert people to the dangers of permitting rats and molluscs to accumulate in the environment, (ii) be diligent supervising children in the backyard and (iii) educate chefs and anyone preparing salads using fresh vegetables about the need to remove slugs and snails and their mucus trails by assiduous washing. From a veterinary perspective, we need more funded research to develop anthelmintic strategies that prevent infection of dogs. Long acting moxidectin formulations given to prevent heartworm seem the most promising line for further studies⁵.

Although to the best of our knowledge neosporosis has never been reported in human patients, recent evidence of infections in sheep²⁷, rhinos²⁸ and carnivorous marsupials²⁹ suggest that a human case will turn up sooner or later! It would be a terrible irony if the patient was a vet.