Members of the genus Hepatozoon have life cycles requiring 2 hosts, an invertebrate definitive host and a vertebrate intermediate host. The definitive host for Hepatozoon americanum, the causative agent of American canine hepatozoonosis (ACH), is the Gulf Coast tick (GCT), Amblyomma maculatum, in which sexual and asexual reproduction results in the production of sporozoite-containing oocysts.1,2 Conventionally, the mode of transmission to canids is assumed to be ingestion of oocysts within H. americanum-infected A. maculatum. Recently, an alternate pathway for transmission of H. americanum (Fig 1) was demonstrated.3–5
Host preference studies suggest that dogs are not favored hosts for A. maculatum.6 Moreover, a natural intermediate host, which serves as a source of infection for the tick definitive host has yet to be identified.2 The majority of reported cases of ACH have occurred in free roaming dogs, whose diet may include prey. Furthermore, a high percentage of coyotes, which survive by predation, are infected with H. americanum.2 Thus, predation of hosts infested with H. americanum-infected A. maculatum ticks is likely an important means of transmission to canids.2
Most confirmed cases of ACH in Oklahoma (>100) have occurred in individual animals as single presentations (RJ Panciera, unpublished data). The purpose of this study is to describe 2 multiple-animal outbreaks of ACH that provided an opportunity to investigate factors involved in the natural transmission of H. americanum to canids, to describe findings that support the importance of predation for canine exposure, and to summarize findings from experimental transmission studies of H. americanum to selected mammalian prey of canids that host immature GCTs.
Case 1: The 1st multiple dog outbreak occurred in a rural mobile home settlement, a site where 2 recently confirmed cases of ACH had occurred and additional dogs were symptomatic. Three dog owners, living in adjacent home sites, were interviewed: 1 had 2 confined dogs; another had 1 confined dog and 1 free-roaming dog; and a 3rd owner, a collector of abandoned animals, had 1 house dog, 6 dogs confined in a well-fenced, mowed, and insecticide-treated backyard, and 2 dogs that roamed freely. The latter owner claimed to have lost 8 dogs, including the 2 confirmed cases of ACH, with similar signs over the past 2–3 years. During the on-site visit, observations of the local environment and the dogs were made and blood samples from 6 of 13 dogs were collected for CBC, blood smear examination, and polymerase chain reaction (PCR) assay of a characteristic fragment of the 18S rRNA gene of Hepatozoon spp. conducted according to the protocol of Allen et al.7A. maculatum, acquired from trapped rodents and from an adjacent pasture, were dissected and examined microscopically for oocysts. GCTs were not found on the dogs.
Neutrophilic leukocytosis was evident in 3 of the 6 dogs bled at the time of the visit. Subsequently, muscle biopsies were collected from 5 of these 6 dogs. Positive PCR assays and “onion skin” cysts present in muscle biopsies from 2 of 3 dogs with leukocytosis confirmed the diagnosis of ACH. Both infected dogs roamed the surrounding area and routinely engaged in predation. Evidence of infection was not observed in the 4 confined dogs. Oocysts were not detected in A. maculatum collected from trapped rodents, the premises, or the adjacent field.
Experimental transmission trials of H. americanum to laboratory-raised cotton rats, mice, and white rats were conducted.3 Cystozoite stages that were infectious for a dog4 were found in multiple tissues of cotton rats and mice. Rats did not become infected.
Case 2: The 2nd multiple dog outbreak of ACH occurred in early January 2008, when a 1-year-old Beagle (dog A), 1 of 6 Beagles owned and hunted by 1 individual, was presented with malaise, a stiff gait, and inappetence. Mild dehydration, lethargy, dull hair coat, tense abdomen, pain on palpation, and rectal temperature of 39.7°C were noted. Five days later dog A and a 2nd ill dog (B) were brought to the clinic. Dog A had a rectal temperature of 40.2°C, persistent malaise and stiffness, mucopurulent ocular discharge, and pronounced neutrophilic leukocytosis (61.3 × 103/μL; reference range, 2.06 × 103–10.6 × 103/μL). Dog B had a rectal temperature of 39.7°C and a neutrophilic leukocytosis (36.3 × 103/μL; reference range, 2.06 × 103–10.6 × 103/μL). The owner reported 2 days later that a 3rd dog (C) was similarly ill. The presumptive diagnosis of ACH was confirmed by muscle biopsy, PCR assay of blood for Hepatozoon spp., blood smear examination, or some combination of these in the 3 dogs. They were treated for ACH as described previously.8
Infection with H. americanum eventually was established in 5 of the 6 Beagles within 13 days of the diagnosis being confirmed in dog A. Single muscle biopsies collected from dogs A, B, and C at 15-day successive intervals demonstrated parasitic cysts of increasing average diameter, indicating that exposure to H. americanum had occurred recently, and probably simultaneously among these dogs.9 However, it was unclear where exposure occurred.
A training exercise, regularly practiced by the owner of dog A, presented an opportunity for simultaneous exposure of the infected group of Beagles to a potential source of infectious material. Dogs were allowed to compete for and consume skin and tissues of single eviscerated rabbits held suspended over the dogs after hunts. Additional blood samples were subsequently collected from 4 nonhunting dogs of the owner and nearby neighbors and from a 2nd group of 5 hunting Beagles, owned and housed separately by another individual. The 2nd group of Beagles almost always accompanied the dogs of the 1st owner in hunts but only sporadically participated in the training exercise of competing for and consuming a rabbit carcass from the hunt. The 2nd owner's dogs had accompanied the 1st owner's dogs on the 1st hunt of the season in late November, the presumed time of exposure. A coyote, retrieved from the 1st owner's property, also was subjected to necropsy and histologic examination. Evidence of Hepatozoon infection was not observed in the 4 nonhunting dogs, in the coyote, or in the 2nd group of hunting Beagles. None of these canids had ingested rabbit carcasses retrieved from the 1st hunt.
Nine rabbits (Sylvilagus spp.) then were collected from the original hunting grounds 8 weeks after canid exposure presumably occurred. The 9 rabbit carcasses were submitted for necropsy and histologic examination. No A. maculatum were found on the rabbit furs. Stages of Hepatozoon were not observed in the tissues of the 9 wild-caught rabbits.
The occurrence of multiple animal outbreaks frequently offers enhanced opportunity to acquire comprehensive information about a given syndrome. In the 1st outbreak, only free-roaming dogs known to actively engage in predation were confirmed to be infected. In the 2nd outbreak, several factors led to the strong belief that a rabbit carcass used for training hunting dogs was the most likely source of infection for the ill dogs.
The high prevalence of infection in dogs (4 of 15) of case 1 from a small, isolated neighborhood implied that sources of H. americanum were readily available. Furthermore, prevalence of infection in the free roaming (4 of 7) rather than confined dogs (0 of 8) supported our suspicion that prey are a major source of infection for dogs and coyotes.
Because small rodents are preferred hosts for A. maculatum6 we conducted both a field survey of small rodents for Hepatozoon infection10 and experimental transmission trials of H. americanum using selected rodent species.3,4Hepatozoon species were present in 18 of 31 wild-trapped cotton rats and 8 of 24 wild-trapped mice with sequences 90.6 and 91.4% identical to reported H. americanum sequences, respectively.10 Laboratory-raised cotton rats and mice fed H. americanum oocysts became infected with cystozoites,3 dormant forms of H. americanum, that were infectious for a dog.4 Rats were not susceptible to infection.
In case 2, evidence of A. maculatum infestation on or near the kennel premises was lacking as was evidence of Hepatozoon infections in nonhunting dogs and the coyote from the 1st owner's kennel premises. Thus, the source of H. americanum for the ill Beagles was unlikely to have come from the kennel and adjacent property. Lack of infection in 5 hunting Beagles separately housed and owned by the 2nd owner but present at the hunt during the presumed time of exposure suggested exposure was unlikely to have occurred at the hunting grounds.
The conviction that 5 of 6 companion Beagles in case 2 acquired H. americanum from a single source is based on (1) the dogs' access to rabbits before the onset of disease was 4–6 weeks, an incubation period similar to that observed in experimentally induced infections1; (2) only dogs that participated in the training exercise of competing for an eviscerated rabbit from the hunt became infected; (3) the progressive increase in the size of parasitic cysts observed in muscle biopsies taken at successive 15-day intervals from the 1st confirmed case indicated that exposure in the 3 dogs biopsied probably occurred simultaneously9; and (4) individually owned and housed Beagles of the 2nd group that had participated in the hunt but not the training exercise were not infected with H. americanum. Collectively, these facts suggested that access to a source of infection, the amount of the infectious material available to dogs at the time of the hunt, or both was probably very limited. The 5 dogs with ACH in case 2 likely were exposed and successfully infected in late November (1st hunt of the season) by a single member of a minimally infectious rabbit population with clinical disease evident in January.
Canine exposure to H. americanum in case 2 could have been from infected A. maculatum on the coat of the ingested rabbit or from cystozoites in the rabbit tissues. The experimental transmission trial established the susceptibility of rabbits to infection with H. americanum and the potential for rabbits to be paratenic hosts harboring cystozoites.5 A retrospective effort to determine the stage of the parasite that was ingested by the dogs in case 2 was not definitive.
The absence of A. maculatum in 9 rabbits retrieved from the hunting grounds in January was not unexpected because rabbits are not usual hosts of A. maculatum and the ticks are active primarily from March through early November in Oklahoma.6 Interestingly, although stages of the organism were not observed microscopically, 2 of 9 wild-caught rabbits appeared to be infected with a Hepatozoon spp. as determined by PCR assay of skeletal muscle. The sequences were 98.8% identical to each other but 92.6 and 93.0% (FJ895406) and 92.4 and 92.8% (FJ895407) identical, respectively, to published sequences of H. americanum from Alabama (AY864676) and Oklahoma (AF176836). Phylogenetic analysis and comparison of the Hepatozoon sequences from the rabbits clustered more closely with sequences reported from various carnivores than sequences reported from other vertebrates (KE Allen et al, unpublished data), suggesting a possible predator–prey relationship for the organism in the rabbit.
It remains important to identify a natural vertebrate reservoir of infection for A. maculatum. Observations made as a result of these 2 ACH outbreaks identified an important alternate mode of transmission to canids and led to follow-up studies that identified 3 potential paratenic hosts for H. americanum: the New Zealand white rabbit (Oryctolagus cuniculus), the cotton rat (Sigmodon hispidus), and mice (Mus musculus). Field observations and experimental findings further verify that predation is important in transmission of H. americanum to dogs, with exposure occurring by ingestion of infected ticks on prey or by ingestion of prey-containing cystozoites.