First Documented Cases of Canine Neuroangiostrongyliasis Due to Angiostrongylus cantonensis in Hawaii
ABSTRACT
Two young dogs domiciled in Honolulu, Hawaii, were presented in November and December 2018 (respectively) for spinal hyperesthesia, hindlimb weakness, and proprioceptive ataxia. Both dogs had neurologic findings referable to spinal cord disease. Both dogs had a combination of lower motor neuron signs (reduced muscle mass, decreased withdrawal reflexes, low tail carriage) and long tract signs (conscious proprioceptive deficits, crossed extensor response, increased myotatic reflexes). Peripheral eosinophilia was present in the second case, but hematology and serum biochemistries were otherwise unremarkable. Plain radiographs and computed tomography scans ± contrast were unremarkable. Cerebrospinal fluid (CSF) from both patients demonstrated eosinophilic pleocytosis, and real-time polymerase chain reaction testing demonstrated Angiostrongylus cantonensis deoxyribonucleic acid in CSF, confirming a diagnosis of neuroangiostrongyliasis. Treatment included glucocorticoid therapy, ± anthelmintic (fenbendazole). Both dogs made a complete recovery. These are the first confirmed cases of autochthonous neuroangiostrongyliasis in canine patients in the United States and the first dogs anywhere to be diagnosed definitively with A cantonensis infection based on real-time polymerase chain reaction testing of CSF. A clinician examining a patient with severe spinal hyperesthesia and a combination of upper and lower motor signs should consider A cantonensis as a differential, especially in endemic areas.
Introduction
The first cases of canine neuroangiostrongyliasis caused by Angiostrongylus cantonensis in Hawaii and the United States are described. The clinical presentation, laboratory findings, treatment, and preventive measures of this emerging infectious zoonotic disease are discussed.
Case Reports
Case 1
In November 2018, a 7 mo old 9.8 kg intact male French bulldog presented to a veterinary referral center in Honolulu, Hawaii, for lethargy, lumbar pain, urinary incontinence, and vomiting, progressing over an 8 day period. The dog had been seen previously at a primary clinic 7 days after clinical signs were first noted but prior to the development of overt neurologic signs, incontinence, and vomiting. Routine hematology and serum biochemistries were un-remarkable at that time.
Upon referral (day 12 of the clinical course), the dog was ambulatory with hindlimb proprioceptive ataxia and decreased hindlimb muscle mass. Conscious proprioception was absent in the right hindlimb and slow in the left hindlimb, while withdrawal reflexes were reduced in both hindlimbs. The dog was hyperesthetic and reacted strongly to palpation of the lumbar spine and penis. Mentation, cranial nerve examination, and the rest of the neurologic examination were unremarkable. Radiographs showed hemivertebrae of T9 with narrowing of the T8–9 and T9–10 intervertebral disc spaces, common abnormalities in this breed. Gabapentin (10 mg/kg per os [PO] q 8 hr) and tramadol (5 mg/kg PO q 8 hr) were prescribed to provide analgesia.
The following day (day 13), the dog was still ambulatory with paraparesis and proprioceptive ataxia. Conscious proprioception was delayed in both hindlimbs, withdrawal reflexes were decreased in both hindlimbs, a crossed-extensor reflex was present, patella and gastrocnemius tendon reflexes were increased bilaterally, and the cutaneous trunci response was lost at the level of L3.
Computed tomography (CT) scans of the head, neck, thorax, abdomen, and lumbosacral regions, before and after IV iohexol (iodinated contrast medium), showed no significant lesions. Cerebrospinal fluid (CSF) collected from the L5–6 intervertebral space was grossly cloudy, and fluid analysis showed 4050 cells per microliter (reference interval 0–5 cells per microliter), protein 209 mg/dL (reference interval ≤45 mg/dL), and glucose 5 mg/dL (reference interval not provided), with marked eosinophilic pleocytosis (eosinophils 88%, small lymphocytes 7%, neutrophils 4%, and macrophages 1%; Figure 1). A diagnosis of eosinophilic meningomyelitis was made. The differential diagnoses included parasitic (e.g., toxoplasmosis, neosporosis, migrating nematodes such as A cantonensis), fungal/algal (e.g., cryptococcosis, protothecosis), idiopathic inflammatory (eosinophilic meningomyelitis, granulomatous encephalomyelitis, steroid-responsive meningomyelitis), viral (e.g., canine distemper virus), and, less likely, paraneoplastic processes.1 Dexamethasone sodium phosphate (SP) was administered IV (0.2 mg/kg), followed by immunosuppressive doses of oral prednisone (1 mg/kg q 12 hr).



Citation: Journal of the American Animal Hospital Association 57, 1; 10.5326/JAAHA-MS-6989
An aliquot of the CSF was sent to the Real-Time PCR Research and Diagnostics Core Facility of University of California Davis (UC Davis) for testing for the following pathogens: canine distemper virus, West Nile virus, Borrelia burgdorferi, Neospora hughesi and caninum, Toxoplasma gondii, Anaplasma phagocytophilum, Ehrlichia canis, Rickettsia spp., and A cantonensis. The real-time polymerase chain reaction (qPCR) test was positive for A cantonensis only, with a cycle threshold of 26.23, confirming neuroangiostrongyliasis. The dog had not been on heartworm preventive and was reported to be an indiscriminate eater with environmental exposure to slugs. The puppy was born in Honolulu, Hawaii, and had no travel history off the island of Oahu.
By the next day (day 14), the dog’s appetite and clinical signs had improved with normal conscious proprioception in all limbs and no pain on spinal palpation. Two weeks later, the patient returned for a follow-up visit and had no demonstrable neurologic deficits or pain. Prednisone was tapered over 4 wk and the patient remained clinically normal at the time of writing.
Case 2
One week after the first case had been presented, a 4 mo old 12.7 kg American pit bull terrier mix presented to the same referral center for inappetence of 3 days’ duration and a reduced energy level. This dog was reported to be an indiscriminate eater with environmental exposure to slugs and had been receiving a monthly heartworm preventive containing milbemycin oxime. There was no travel history off the island of Oahu. The dog was trembling and stood with a hunched posture. It was ambulatory, with a short-strided pelvic limb gait and reluctance to move its hindlimbs. Pain responses were evoked by range of motion manipulation of the hips and stifles and palpation of the stifles and thoracolumbar and lumbosacral spine. Pelvic radiographs were unremarkable.
Complete blood cell count showed leukocytosis (27,500 cells per microliter), neutrophilia (15,720 cells per microliter), lymphocytosis (5750 cells per microliter), eosinophilia (4230 cells per microliter), and basophilia (410 cells per microliter). No abnormalities were detected in routine serum chemistry tests. The owner had observed the dog passing a large quantity of worms in its stool 3 days earlier. The dog was treated empirically using amoxicillin clavulanate (25 mg/kg PO q 12 hr), pyrantel pamoate (5 mg/kg), tramadol (4 mg/kg PO q 6 hr), and gabapentin (8 mg/kg PO q 8 hr).
The following day (day 4), the patient was nonambulatory, with knuckling of both hindlimbs, and was mentally dull. Neurologic examination revealed absent conscious proprioception in both hindlimbs, exaggerated patellar reflexes, and severe cervical, paraspinal, and tail base hyperesthesia. Mentation, cranial nerve examination, and the rest of the neurologic examination were unremarkable. CT scans of the neck, thorax, abdomen, and lumbosacral regions, before and after IV iohexol, were unremarkable. CSF collected from the L5–6 intervertebral space contained 6000 cells per microliter (reference interval 0–5 cells per microliter), protein 272 mg/dL (reference interval ≤45 mg/dL), glucose 23 mg/dL (reference interval not provided), and severe eosinophilic pleocytosis (eosinophils 91%, small lymphocytes 7%, macrophages 2%, and neutrophils <1%). A diagnosis of eosinophilic meningoencephalitis was made with the same differential diagnoses as Case 1, and dexamethasone SP was administered IV (0.2 mg/kg), followed by prednisone (0.8 mg/kg PO q 12 hr). Tramadol and gabapentin were continued. By that evening, the dog appeared more comfortable and was eating well. The next day (day 5), the patient remained nonambulatory with significantly improved motor function. The neurologic exam remained unchanged, but the dog was less hyperesthetic. Doxycycline (6.25 mg/kg PO q 12 hr) and clindamycin (12.5 mg/kg PO q 12 hr) were added to the drug regimen and the prednisone dosage was increased to 1 mg/kg PO q 12 hr.
An aliquot of CSF was sent to UC Davis for qPCR testing for the same neuropathogens as per Case 1: the qPCR test was positive for A cantonensis alone, with a cycle threshold of 35.24, confirming a diagnosis of neuroangiostrongyliasis.
By day 7, the dog was able to walk unassisted but remained paraparetic with proprioceptive ataxia. At this stage, conscious proprioception was present in all four limbs. Clindamycin was discontinued, and fenbendazole was prescribed (25 mg/kg PO q 24 hr for 21 days). Doxycycline was continued for 3 wk and prednisone was tapered over that time period. Seven days after discharge, the patient returned for a follow-up visit and all clinical signs had resolved. The dog remains clinically normal at the time of writing.
Discussion
The rat lungworm A cantonensis is a parasitic nematode transmitted between rats (the definitive host) and slugs or snails (intermediate hosts) in its natural life cycle.2 Adult male and female worms reside in the right ventricle and pulmonary arteries and lay eggs, which embolize to the lung where they develop. First-stage larvae (L1)are coughed up and swallowed, pass through the digestive tract, and are shed in feces. The L1 larvae are ingested by intermediate mollusk hosts and develop into infective third-stage larvae (L3) while preserving a double cuticle. Rats are infected by ingesting intermediate hosts harboring L3 larvae. These L3 larvae penetrate the rats’ intestinal walls, enter the portal circulation, and migrate through the liver, heart, lungs, and kidneys before entering the central nervous system (CNS). In the spinal cord and brain, L3 larvae feed, burrow, grow, molt (shedding their double cuticles), and eventually mature into subadults, which penetrate the meningeal veins to travel to the right ventricle and pulmonary arteries where they mature and reproduce.3
Humans and canines are accidental “dead-end” hosts that typically acquire the infection via deliberate or inadvertent ingestion of infective L3 larvae intermediate or paratenic (e.g., freshwater crustaceans, frogs, toads, lizards, centipedes, etc.) hosts. After initial infection, the parasite life cycle progresses in the same way as in definitive hosts but generally only to the subadult stage. Sub-adults are in general unable to leave the CNS, where they cause granulomatous/eosinophilic meningitis and encephalitis resulting in variably severe CNS disease (rat lungworm disease) in accidental hosts.3 The parasite was historically recognized in Southeast Asia and the Pacific Islands, although its distribution has expanded and now includes Australia, Africa, the Caribbean, and the mainland United States.2,3 Of recorded human cases from the United States, the majority have occurred in patients living in or visiting Hawaii.4
These two case studies represent the first confirmed cases of autochthonous canine neuroangiostrongyliasis caused by A cantonensis infections in Hawaii, and indeed the United States. These reports also represent the first two canine A cantonensis infections diagnosed definitively using qPCR on CSF specimens. Anecdotally, canine neuroangiostrongyliasis is well recognized as a common clinical entity in dogs domiciled on the Big Island of Hawaii, in Hilo and especially the Puna district (Alfred Mina, personal communication, January 2020). Dogs typically present in a similar manner to the two cases of the current report, and cases are treated along similar lines to the present cases. Confirmation of the diagnosis by collection of CSF is, however, not commonly performed because of financial constraints, and so definitive diagnosis by demonstrating eosinophilic pleocytosis and A cantonensis deoxyribonucleic acid in CSF using qPCR has not been recorded in the peer-reviewed literature. Neuroangiostrongyliasis has also been diagnosed histologically following necropsy in various exhibited animals within a Hilo zoo (Pam Mizuno, personal communication, March 2019). More recently, neuroangiostrongyliasis has been diagnosed in horses on the Big Island (Susan Jarvi, personal communication, January 2020).
The clinical presentation of the two dogs in this report is consistent with the syndromic presentations described by Jindrak and Alicata in experimental dogs and in naturally infected puppies as confirmed by Mason, Collins, Lunn, and their respective colleagues.5–9 The two affected dogs were young, being 7 and 4 mo of age at time of diagnosis, respectively, and presented with lumbar hyperesthesia and hindlimb proprioceptive ataxia and weakness, with Case 1 also having urinary incontinence. Peripheral eosinophilia was detected in Case 2 only, and no abnormalities in the serum biochemistry panel were detected in either dog. Both dogs were initially treated symptomatically for neuropathic pain using tramadol and gabapentin and were further investigated using diagnostic imaging. Radiographs and CT (with and without iodinated contrast) of the spine showed no neural or musculoskeletal abnormalities. MRI would have provided additional information about the underlying distribution of neural lesions but was not readily available. CSF from both dogs showed marked eosinophilic pleocytosis, decreased glucose, and increased protein concentrations, consistent with what is seen in human and previous canine cases.10–13 The combination of characteristic clinical signs including lumbar hyperesthesia and eosinophilic pleocytosis of CSF led to a suspicion of neuroangiostrongyliasis. Accordingly, corticosteroid therapy was started, initially using IV dexamethasone SP and subsequently oral prednisone, pending multiplex PCR testing for canine neuropathogens. A cantonensis deoxyribonucleic acid was detected by qPCR in both patients, confirming the presumptive diagnosis, whereas other potential pathogens were considered less likely based on negative qPCR results. Both dogs showed a beneficial response to corticosteroids within 24–48 hr. Dog 1 did not receive anthelmintics owing to profound improvement with corticosteroids alone; dog 2 was treated additionally with fenbendazole (25 mg/kg orally once daily) for 3 wk, in accordance with the prevailing trends in management of human cases.10 Upon recheck 10–14 days later, clinical signs in both dogs had resolved completely.
A variety of treatment approaches have been described in the literature, and to date there have been no prospective studies carried out to determine the best treatment regimen for canine patients with A cantonensis–induced neuroangiostrongyliasis.6–9 The current rationale is based on the understanding that damage to host neural tissues is due in part to mechanical injury from migrating L3 larvae and in part to the eosinophilic inflammatory response to nematode antigens released during growth and especially molting.14 Corticosteroids are used to dampen down the immune response to foreign antigens released by the migrating parasites, and subsequently, fenbendazole is used to gradually kill the L3 larvae, thereby preventing further mechanical damage to the neural tissues due to migration and growth of the parasites.10–14 Although older articles are hesitant to recommend a regimen in which anthelmintics are used to kill the migrating larvae, more recent case reports and case series generally advocate the bipartite approach.6,9–13 Doxycycline, trimethoprimsulfonamide, and clindamycin are often used empirically to treat progressive multifocal encephalomyelitis until a definitive diagnosis can be made and to cover the possibility of enteric bacteria being translocated by the migrating nematode larvae (Trojan horse phenomenon).9
Previous case reports describe variable response to treatment of A cantonensis infection in canine patients, with some dogs recovering completely, some having only partial response to treatment or a prolonged recovery, and some failing to respond to treatment with corticosteroids and anthelmintic medications entirely.6–9 Dogs with more severe neurologic deficits may have poorer outcomes.6–9 Both dogs in this report were diagnosed and treatedearly inthecourseofthediseaseprocess,shortlyafteronset of neurologic deficits, which likely played a role in the positive outcome in these two cases.
The two cases documented in this report help to establish A cantonensis as an emerging and important canine infectious disease in the Hawaiian Islands. Marked lumbar and tail hyperesthesia and paresis primarily affecting the pelvic limbs are commonly described clinical signs that may help distinguish neuroangiostrongyliasis from other causes of meningitis. Unlike many other causes of paraparesis in dogs, neuroangiostrongyliasis can present with a combination of signs of upper motor neuron disease (crossed extensor response, increased myotatic reflexes, conscious proprioceptive deficits) and lower motor neuron disease (reduced muscle mass, decreased withdrawal reflexes, reduced tail wag, low tail carriage). Signs are presumably caused by migrating L3 larvae that initially are present in large numbers in the cauda equina and move rostrally, causing a wave of damage to the spinal cord parenchyma. Neuroangiostrongyliasis should therefore be considered in canine patients with severe spinal hyperesthesia that is out of proportion with the extent of paraparesis and a combination of upper and lower motor neuron signs, especially in endemic areas such as the Hawaiian Islands.15 Given the rapid and complete response to therapy in these dogs and importance of early intervention, empiric treatment with steroids and anthelmintics is warranted in cases with consistent clinical signs where a definitive diagnosis is not possible. In areas where A cantonensis is endemic, routine administration of a monthly commercial transdermal moxidectin product to prevent accidental acquisition of neuroangiostrongyliasis should be considered. Prophylactic administration of a 5 day course of fenbendazole or application of transdermal moxidectin following ingestion of a slug or snail may be considered as prophylaxis to prevent migrating larvae from reaching the CNSa, b. Further studies are necessary to determine the efficacy of preventive and prophylactic treatment strategies. Additionally, owners should be educated about the parasite’s life cycle and, depending on the likely routes of exposure, consider rat and/or mollusk control to prevent exposure.
Dogs are an important beacon or sentinel for determining human risk of A cantonensis exposure in a given geographic region as they are less likely to travel and tend to be indiscriminate eaters with greater exposure to intermediate hosts. The authors recommend a one health approach whereby both human and canine cases of A cantonensis infection are reported to better establish geographic areas where infection is possible and help increase awareness and ability to implement preventive strategies where they are most needed.

Photomicrograph of cytospin preparation of cerebrospinal fluid from Case 1 demonstrating eosinophilic pleocytosis. Original magnification ×60 (Wright’s stain; bar = 20 microns).
Contributor Notes


