Pufferfish exposure and presumed tetrodotoxicosis in Australian dogs: 372 cases (2007–2018)

Abstract

Introduction

Tetrodotoxin (TTX) is a neurotoxin found in pufferfish that results in the blockade of excitatory conduction and clinical signs including gastrointestinal upset and paralysis. The objectives of this study were to describe the epidemiology of pufferfish exposure and presumed tetrodotoxicosis in dogs, and to determine associations with the development of tetrodotoxicosis.

Methods

This was a multicentre historical cohort study with cases enrolled from veterinary hospitals in Perth Western Australia, and southeast Queensland. A medical record search identified cases between January 1^st 2007 and 5^th October 2018. Dog signalment, details of pufferfish exposure and veterinary presentation, clinical signs, treatment and outcome were recorded. Chi square tests were used to investigate the association between pufferfish exposure features and development of presumed tetrodotoxicosis.

Results

Three hundred and seventy‐two dogs were included. Two hundred and seventeen had pufferfish exposure only, while 155 dogs developed presumed tetrodotoxicosis. Gastrointestinal signs were the most common clinical signs prior to presentation (149/372, 40.1%), followed by neuromuscular signs (40, 10.8%). Dogs were significantly more likely to show clinical signs of tetrodotoxicosis if they ingested more than one pufferfish and had a delay to presentation of more than 2 h. Most dogs had gastrointestinal decontamination and were treated as outpatients (292, 78.8%). Eight received mechanical ventilation. All but 4 dogs survived (369/372); two dogs died naturally and two were euthanised.

Conclusions

Clinical signs are generally mild when presumed tetrodotoxicosis occurs after pufferfish ingestion in dogs. The prognosis for complete recovery is good with timely initiation of decontamination and supportive care.

Abbreviations

ADME

absorption, distribution, metabolism, and excretion

AES

Animal Emergency Service

CDI

central diabetes insipidus

Cmax

peak plasma concentration

CPR

cardiopulmonary resuscitation

CRI

continuous rate infusion

CRTZ

chemoreceptor trigger zone

EMR

electronic medical records

ICU

intensive care unit

IVLE

intravenous lipid emulsion

LD50

median lethal dose

PVE

perth vet emergency

REDCap

research electronic data capture

ROSC

return of spontaneous circulation

TAHMU

The Animal Hospital at Murdoch University

TTX

tetrodotoxin

USG

urine specific gravity

WA

western Australia

Tetrodotoxin (TTX) is a neurotoxin that binds to voltage‐gated sodium channels preventing the influx of sodium ions during action potentials, which results in the blockade of excitatory conduction in most neurons and myocytes. ¹ This toxin was first isolated from the ovaries of a globefish, and then named tetrodotoxin after the taxonomic family of pufferfish from which it was isolated, Tetraodontidae. ² It has since been detected in many phylogenetically unrelated marine organisms likely as a consequence of its production in symbiotic bacteria. ³ , ⁴ , ⁵

Marine pufferfish have caused many incidents of tetrodotoxicosis, especially in Japan where pufferfish, or ‘Fugu’ is considered a delicacy. ⁴ Indeed several people still die each year due to its ingestion. ⁴ As such, there is considerable literature available on the effects of TTX in humans. ⁴ , ⁶ , ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹ Clinical signs in humans affect multiple body systems including the neuromuscular, gastrointestinal, cardiovascular and respiratory systems. A four‐step grading scheme has been described, with some variation between authors, where grade one is the most mild and grade four is the most severe clinical signs, ² , ⁵ , ¹² as outlined in Table 1. There is no antidote, thus management involves decontamination and supportive care until the toxin is eliminated from the body.

Table 1.

Clinical signs (and symptoms) affecting different body systems associated with grades 1–4 tetrodotoxicosis in humans

Grade	Gastrointestinal	Neuromuscular	Respiratory and cardiac
1	Vomiting, nausea Hypersalivation Hematemesis Hypermotility Diarrhoea Abdominal pain	Perioral paresthesia Headache Diaphoresis Pupillary constriction
2		Paresthesia spreading to the trunk and extremities Early motor paralysis Lack of coordination Slurred speech
3		General flaccid paralysis with reduced muscle tone Aphonia Fixed or dilated pupils
4			Respiratory paralysis Hypoxia Hypotension Cardiac dysrhythmias Impaired consciousness

The pharmacokinetics of TTX have been predominantly studied in rodents, with incomplete investigation in dogs. Indeed, studying the absorption, distribution, metabolism and excretion (ADME) of TTX is reportedly difficult due to the very low lethal dose. ¹³ The median lethal dose (LD₅₀) in rats is approximately 10 μg/kg administered by intramuscular (IM) injection, ¹³ while another study reported an LD₅₀ of 2.85 μg/kg administered orally (by feeding) in mice. ¹⁴ Similarly, 10 μg/kg administered subcutaneously and 4.8 μg/kg intravenously are described as lethal doses in dogs, ¹⁵ , ¹⁶ while 70 μg/kg per os reliably causes vomiting but is not lethal in dogs. ¹⁵ Although the amount of TTX in a single pufferfish is extremely varied, there is evidence that a single pufferfish contains several to tens of milligrams of TTX, ⁴ and thus is enough to result in lethality in a human or dog. In a study investigating the ADME of 6 μg/kg radiolabelled TTX IM in rats, peak plasma concentration (C_max) was reached in 10 min, the mean residence time was 1.62 h, and plasma half‐life was 2.31 h, with the majority of the toxin excreted unchanged in the urine, and only negligible biliary excretion. Distribution was highest in the stomach wall, kidney, lungs, heart and intestine, with the average total radioactivity in these organs being higher than or equivalent to plamsa. ¹³ This pattern of distribution is consistent with its low molecular weight and low plasma protein binding, suggesting a large volume of distribution. Studies of oral administration in rodents demonstrate rapid absorption with peak plasma concentrations reached in the first hour after feeding TTX. ¹⁷

Literature on tetrodotoxicosis in dogs is limited to a single publication describing neurotoxicosis after the ingestion of the grey side‐gilled sea slug in New Zealand, ¹⁸ despite a plethora of experimental literature regarding the effects of TTX on canine cells (generally ex vivo). ¹⁹ , ²⁰ , ²¹ The staff at The Animal Hospital at Murdoch University (TAHMU) in Perth, Western Australia (WA), can attest to the frequency of pufferfish ingestion in local dogs, and this experience is shared by other veterinarians in WA ²² and eastern Australia. ²³ Anecdotal experience is that many dogs that ingest pufferfish vomit them up (self‐decontamination) and are otherwise well, or have emesis induced at a veterinary clinic. Rare cases are anecdotally reported of dogs requiring mechanical ventilation for tetrodotoxicosis. Although tetrodotoxicosis is mentioned in veterinary texts and review articles, ²⁴ there is no published literature documenting the epidemiology of pufferfish exposure and presumed tetrodotoxicosis in dogs. This study aimed to fill this knowledge gap.

The primary objective of this study was to describe the epidemiology of pufferfish exposure and presumed tetrodotoxicosis in Australian dogs. A secondary objective was to determine if the number of pufferfish ingested or the time to presentation to a veterinary hospital was associated with the likelihood of tetrodotoxicosis developing after pufferfish exposure. Based on our anecdotal experience, we hypothesised that when tetrodotoxicosis occurs after pufferfish ingestion, clinical signs are mild in the majority of dogs with a good prognosis for complete recovery with supportive care. We also hypothesised that dogs would be more likely to develop clinical signs of presumed tetrodotoxicosis if they ingested multiple pufferfish, or presentation to a veterinary hospital was delayed.

Materials and methods

This study was designed as a multicentre historical cohort study. Participating centres included TAHMU in WA, as well as three hospitals from the Animal Emergency Australia group; Perth Veterinary Emergency (PVE) in WA, and two Animal Emergency Service (AES) hospitals in southeast Queensland (Underwood, and Carrara). Owner consent was provided for treatment and hospital admission, which includes consent for patient data use in research and publication as a clause in the consent form. The electronic medical records were searched using search terms specific to pufferfish exposure and tetrodotoxicosis. Search terms included ‘puffer fish’, ‘pufferfish, ‘blow fish’, ‘blowfish’, ‘tetrodotoxin’ and ‘tetrodotoxicity’. The period covered by the search was January 1^st 2007 (the first date on which the searchable TAHMU EMR was in use) to the date when the search was conducted (5^th October 2018). Cases from the Animal Emergency Australia group were included from the same time period. Medical records were then individually screened, with confirmed pufferfish ingestion in dogs being the inclusion criteria. Dogs were also included if they had been seen licking a pufferfish or with a pufferfish in their mouth, even if it was not swallowed. Since pufferfish have a characteristic appearance and are commonly found discarded by fishermen on ocean beaches or by inland saltwater and brackish waterways, owner‐witnessed ingestion or owner or veterinarian identification of pufferfish in the vomitus was considered confirmatory. Dogs were classified as exposed only if they ingested pufferfish or had a pufferfish in their mouth but did not develop clinical signs. Dogs were classified as experiencing presumed tetrodotoxicosis if they had confirmed ingestion and developed consistent clinical signs, either before or after presentation to a veterinary hospital. The term ‘presumed tetrodotoxicosis’ is used in this case series since toxicology testing was not performed to confirm the presence of TTX in the affected dogs. Dogs that had compatible clinical signs and possible, but not confirmed exposure were excluded, even if tetrodotoxicosis was the presumptive diagnosis documented in the medical record.

Medical records were reviewed and patient information was recorded using a standardised data collection instrument in the Research Electronic Data Capture (RedCAP) platform; a secure, customisable, online database for medical research. ²⁵ , ²⁶ To minimise data entry errors two authors (JH, CS) were involved in data extraction from the medical records. Either the first author conducted initial data entry which was checked by the second author, or both authors entered data together. The signalment (date of birth, sex, breed), date of pufferfish ingestion, location of ingestion (beach or inland waterway/river), number of puffer fish ingested (if known), presenting vital signs (heart rate, temperature and blood pressure), presenting clinical signs, any decontamination performed (e.g. whether emesis was performed and in what time frame following ingestion), treatment administered and outcome were recorded. Dogs were only considered to have vomiting as a clinical sign if they vomited spontaneously, unassociated with an emetic agent given by the owner, referring veterinarian or at a study hospital. The difference between the date of birth and the date of pufferfish ingestion was used to calculate the patient age within REDCap, in months. Dogs listed as a cross of one or more breeds (e.g. poodle cross) were recorded as mixed breed dogs. Oscillometric mean arterial pressure or Doppler blood pressure (in mmHg) was recorded. Outcome was classified as inpatient or outpatient treatment; for those treated as inpatients, it was further classified as survived to discharge, euthanised or died naturally.

Nonparametric descriptive statistics were used to describe the population. Continuous data is summarised as median (Q1–Q3, Min‐Max), and categorical data is summarised as proportions and percentages. To investigate the association between pufferfish exposure features and development of presumed tetrodotoxicosis, the relevant data from REDCap was exported to R to create 2‐by‐2 contingency tables. Features of pufferfish exposure (e.g. 1 pufferfish vs. >1 pufferfish ingested) were considered the exposure variables, and the development of clinical signs of tetrodotoxicosis was considered the outcome variable. A Chi‐Square test with Yates' continuity correction or Fisher's exact test was performed as appropriate to determine if there is a relationship between the exposure and outcome variables with P < 0.05 considered significant.

Results

The medical records search yielded 749 results, of which 372 were included (Figure 1, Table 2). Two hundred and seventeen had pufferfish exposure only, and did not develop clinical signs at any time, while 155 dogs developed clinical signs consistent with tetrodotoxicosis.

Figure 1.

Results of electronic medical records search for dogs with pufferfish ingestion and/or tetrodotoxicosis at The Animal Hospital at Murdoch University and Animal Emergency Australia hospitals.

Table 2.

Summary statistics of 372 dogs evaluated for pufferfish ingestion and/or presumed tetrodotoxicosis at The Animal Hospital at Murdoch University (TAHMU), Animal Emergency Service (AES), and Perth Vet Emergency (PVE) from 2007 to 2018

	Summary statistics
Parameter	All dogs (n = 372)	Exposure only (n = 217)	Tetrodotoxicosis (n = 155)
Treating hospital
TAHMU	311, 83.6%	202, 93.1%	109, 70.3%
AES	52, 13.9%	9, 4.1%	43, 27.7%
PVE	9, 2.4%	6, 2.8%	3, 1.9%
Age (months)
Median	24	18	36
Q1‐Q3	10–54.5	10–48	13.25–72
Min‐max	3–180	3–180	3–168
Sex
Male neutered	148, 39.8%	87, 40.1%	61, 39.4%
Female spayed	137, 36.8%	76, 35.0%	61, 39.4%
Male intact	39, 10.5%	26, 12.0%	13, 8.4%
Female intact	48, 12.9%	28, 12.9%	20, 12.9%
Breeds (represented by 10 or more dogs)
Mixed breed	89, 23.9%	37, 17.1%	52, 33.5%
Labrador retriever	87, 23.4%	61, 28.1%	26, 16.8%
Staffordshire bull terrier	31, 8.3%	22, 10.1%	9, 5.8%
Golden retriever	19, 5.1%	12, 5.5%	7, 4.5%
Border collie	19, 5.1%	10, 4.6%	9, 5.8%
Other breeds (n = 54, < 10 each)	127, 34.1%	75, 34.6%	52, 33.5%
Body weight (kg)	n = 259	n = 140	n = 119
Median	23	24.7	22.0
Q1‐Q3	16.43–30.58	18.6–30.85	13.15–30.4
Min‐max	2.8–71	5.2–53	2.8–71
Location of pufferfish exposure
Beach	163, 43.7%	87, 40.1%	76, 49.0%
Inland waterway/river	78, 20.7%	46, 21.2%	31, 20.0%
Not documented or unclear	132, 35.4%	84, 38.7%	48, 31.0%
Primary accession vs. referral
Primary accession	354, 95.2%	215, 99.1%	139, 89.7%
Referral	18, 4.8%	2, 0.9%	16, 10.3%
Number of pufferfish ingested
One	309, 83.1%	196, 90.3%	114, 73.5%
Two	22, 5.9%	13, 6.0%	8, 5.2%
Three	6, 1.6%	1, 0.5%	5, 3.2%
More than three	11, 3.0%	3, 1.4%	8, 5.2%
Not reported	12, 3.2%	2, 0.9%	10, 6.5%
Only in mouth, chewed or licked	12, 3.2%	2, 0.9%	10, 6.5%
Number of pufferfish ingested (categorical/binary)—For chi square test
One ingested/in mouth/licked	312	198	123
> One ingested	51	19	32
Time to presentation
Brought in immediately (< 1 h)	175, 47.0%	140, 64.5%	35, 22.6%
1–2 h	85, 22.8%	56, 25.8%	29, 18.7%
2‐3 h	38, 10.2%	14, 6.5%	24, 15.5%
3‐4 h	19, 5.1%	2, 0.9%	17, 11.0%
>4h	39, 10.5%	3, 1.4%	36, 23. 2%
Not reported	16, 4.3%	2, 0.9%	14, 9.0%
Time to presentation (categorical / binary) – For chi square test
Within 2 h	260	196	64
> 2h or not reported	112	21	91
Presenting heart rate (beats/min)	n = 327	n = 184	n = 143
Median	120	120	120
Q1‐Q3	13–140	100–136.5	107.5–140
Min‐max	64–220	70–220	64–200
Presenting temperature (°C)	n = 286	n = 155	n = 131
Median	38.7	38.9	38.5
Q1‐Q3	38.33–39.0	38.5–39.1	38.1–38.8
Min‐max	33.7–40.0	37.3–40	33.7–39.8
Blood pressure (mmHg)	n = 55	n = 16	n = 39
Median	107	98.5	110
Q1‐Q3	96.5–119	90–106.5	100.5–121
Min‐max	70–180	82–156	70–180
Length of hospital stay (days)		n = 6	n = 68
Median	0	1	1
Q1‐Q3	0–1	0.25–1	0–1
Min‐max	0–2	0–1	0–2

Signalment

Included dogs were generally young (median 24 months), with fairly even sex distribution (Table 2). Mixed breeds accounted for 23.9% of dogs, with the Labrador retriever (23.4%) being the most common purebreed, and a variety of other breeds represented (Table 1).

Details of pufferfish exposure and veterinary presentation

When recorded, the location of pufferfish exposure was predominantely the beach (Table 1). Most dogs ingested a single pufferfish (83.1%)(Table 1). The maximum number of pufferfish ingested was 12, which occurred in two separate Labrador retrievers. Some dogs did not actually ingest a pufferfish (n = 12/372, 3.2%) but were rather noted have only to have licked a pufferfish or had a pufferfish in their mouth, with or without chewing it, before dropping it or their owners removing it. Seventy percent of dogs presented within 2 h of ingestion, many of which were brought immediately to the veterinary hospital after witnessed ingestion (Table 1). Three hundred and fifty‐four dogs (95.2%) were primary accession cases, while 18 (4.8%) were referred from other veterinarians.

Dogs referred from other veterinary hospitals

Dogs with clinical signs were referred more commonly (n = 16) than those without (n = 2). Four dogs were referred immediately without any treatment, whereas 14 had some veterinary intervention before referral. Induction of emesis was most common intervention (n = 7), successfully producing pufferfish in the vomit in 4 cases, but without pufferfish in the vomit in 7 cases. Metoclopramide and activated charcoal were also given in 7 dogs. Four dogs were intubated and manually ventilated prior to referral, and manual ventilation continued during transport to a study centre (2 to TAHMU, 2 to AES). Three of these cases were transitioned to mechanical ventilation upon arrival at the referral centre and ultimately survived to discharge, whereas the other did not require ongoing ventilation and was transitioned to nasal oxygen following arrival. Gastric lavage under anaesthesia and administration of intravenous lipid emulsion (IVLE) were performed as additional decontamination in one of the referred ventilated cases. Another of the ventilated cases was treated with subcutaneous doses of dexamethasone, chlorpheniramine and adrenaline prior to referral due to concern for anaphylaxis since he had facial angioedema and moderate‐to‐severe stridor after vomiting a pufferfish.

Clinical signs prior to presentation

The majority of dogs, 58.6% (218/372), were asymptomatic prior to presentation to a veterinarian. One hundred and fifty‐four (154/372 = 41.4%) dogs had clinical signs of tetrodotoxicosis prior to presentation, and some of these dogs had multiple clinical signs. One additional dog that was asymptomatic prior to presentation developed clinical signs following presentation, yielding a total of 155 dogs with clinical signs of presumed tetrodotoxicosis. Table 3 displays the numbers of dogs experiencing different clinical signs both prior to presentation and in hospital.

Table 3.

Clinical signs reported amongst 155 dogs with presumed tetrodotoxicosis prior to and after presentation to The Animal Hospital at Murdoch University, Animal Emergency Service and Perth Vet Emergency

	Number and proportion of dogs displaying clinical signs of n = 155 dogs with presumed tetrodotoxicosis
Clinical signs	Prior to presentation	At presentation/in hospital
None	1, 0.6%	80, 51.6%
Gastrointestinal signs	149, 96.1%	37, 23.9%
Vomiting and/or regurgitation containing pufferfish	88, 56.8%	2, 1.3%
Vomiting and/or regurgitation not containing pufferfish	90, 58.1%	14, 9.0%
Ptyalism	21, 13.5%	16, 10.3%
Diarrhoea	7, 4.5%	3, 1.9%
Evidence of abdominal pain	1, 0.6%	8, 5.2%
Abdominal distention	1, 0.6%	0
Neuromuscular signs	40, 25.8%	41, 26.5%
Weakness or wobbliness (ataxia)	27, 17.4%	25, 16.1%
Collapse, non‐ambulatory (tetraparesis)	13, 8.4%	13, 8.4%
Muscle tremors/fasciculations	7, 4.5%	7, 4.5%
Reluctance to stand	1, 0.6%	0
Lack of gag reflex	N/A	4, 2.6%
Inappropriate mentation	24, 15.5%	9, 5.8%
Respiratory distress	11, 7.1%	16, 10.3%
Cardiovascular signs	N/A	9, 5.8%
Facial angioedema	2, 1.3%	0
Cardiopulmonary arrest	2, 1.3%	1, 0.6%

Percentages displayed are of 155 dogs with clinical signs.

Gastrointestinal signs were the most common clinical signs prior to presentation (149/372, 40.1% of all dogs), followed by neuromuscular signs (40, 10.8%), being mentally inappropriate or obtunded (24, 6.5%), respiratory signs (11, 3%) and facial angioedema (2, 0.5%). One dog experienced cardiopulmonary arrest prior to presentation; more detail is provided regarding this case in the outcome section. Of the dogs with gastrointestinal signs, vomiting and/or regurgitation—either with or without pufferfish—was most common (Table 2). Although the number of times dogs vomited was not recorded in the data collection instrument, medical records commonly described profuse vomiting that continued even after a pufferfish was vomited up. Of those dogs with neuromuscular signs prior to presentation weakness or wobbliness was most common, followed by collapse and muscle tremors or fasciculations.

Physical examination findings and clinical signs after presentation

Presenting heart rate, temperature and blood pressure data were similar among groups (Table 1). Most dogs (297/372, 78.9%) displayed no clinical signs after presentation to a study hospital. Neuromuscular signs were the most common signs of tetrodotoxicosis after presentation and were seen in 11% of all dogs (41 cases); ataxia/weakness/wobbliness (n = 25) and tetraparesis and collapse (13) were the most common. Additionally, 9 dogs were considered mentally inappropriate or obtunded after presentation. Level of consciousness was not able to be accurately assessed in 3 cases given that they had received apomorphine prior to presentation.

Gastrointestinal signs were the second most common clinical signs after presentation (36, 9.7%), with residual vomiting, ptyalism and abdominal pain predominating (Table 3). A small number of dogs (16/372, 4.3%) displayed respiratory signs, including tachypnoea (3), respiratory distress requiring intubation and manual ventilation (5) and respiratory paralysis requiring mechanical ventilation (8). Only 9 dogs (9/372, 2.4%) had cardiovascular signs, including 7 with tachycardia, one with bradycardia and one with hypotension.

Only one dog was reported to have a seizure after presentation, which was not thought to be related to tetrodotoxicosis but rather to an underlying seizure disorder. This was a 10‐year‐old, neutered male Golden retriever with a history of infrequent generalised seizures (approximately two per year). The dog had not undergone any investigation and did not receive treatment for the seizures. This dog had ingested a pufferfish 1–2 h prior to presentation, was clinically normal prior to presentation, and had a normal physical examination at the time of presentation. Emesis was induced with apomorphine (0.04 mg/kg IV once), resulting in vomiting that included a pufferfish and a large volume of bile‐stained malodorous fluid. After repeated vomits, the dog that had a generalised seizure which stopped before diazepam could be administered IV. The dog recovered quickly and was discharged shortly thereafter at the owners' request.

Grade of clinical signs

The grade of clinical signs was determined for 152 cases. Ninety‐eight dogs (64.5%) had grade one clinical signs, 11 (7.2%) were grade two, 30 (19.7%) were grade three, and 13 (8.6%) were grade four. Grade was not determined for three cases since they had received apomorphine at their referring veterinarians, which was suspected to confound their neurological examination.

Associations with tetrodotoxicosis

Dogs that had ingested more than one pufferfish were significantly more likely to show clinical signs of tetrodotoxicosis (32/51, 63%) than those that had only ingested a single pufferfish or only had a licked a pufferfish or had it in their mouth (123/312, 39%), X ² (1, n = 372) = 12.52, P < 0.001. Dogs in which the delay to presentation to a veterinarian was more than 2 h were significantly more likely to show clinical signs of tetrodotoxicosis (91/112, 81%) than those that presented within the first 2 h (64/260, 25%), X ² (1, n = 372) = 103.3, P < 0.00001.

Gastrointestinal decontamination

Emesis was induced in most cases after presentation to a study hospital (270/372, 72.5%). Of the cases that did not have emesis induced at a study hospital, eight had already had emesis induced at their referring veterinarian, and the remainder had already vomited the pufferfish they ingested prior to presentation. Apomorphine was used as the emetic agent in 195 dogs, whereas soda crystals were used in 124 dogs. Some dogs were administered both emetic agents if the first was unsuccessful.

Emesis successfully resulted in the evacuation of pufferfish in 74.4% (201/270) of cases, whereas in 24.5% (66/270) of cases, emesis did not result in the evacuation of a pufferfish or did not induce vomiting at all (2/270, 0.7%). The success of emesis was not reported in two cases. Induction of emesis was still successful in 11 dogs that had prior spontaneous vomiting without producing pufferfish. Four dogs had gastric lavage performed under general anaesthesia and one had endoscopy to remove remaining pufferfish from the stomach. Activated charcoal was also commonly given at study hospitals, with 209/372 dogs (56.2%) receiving a single dose, and 3 receiving multiple doses.

Treatment

A small number of dogs were admitted to hospital for further care/observation (76/372, 20.4%), whereas the vast majority were treated on an outpatient basis and sent home (292, 78.8%), or were deceased or euthanised at presentation (3, 0.8%).

A total of 20.2% of cases (75/371) received no medications, including two that were hospitalised, with their hospitalisation being purely for monitoring (Table 5). Of the those that received treatment, most dogs were treated with metoclopramide as an antiemetic (Table 5), particularly after emesis was induced. Other antiemetic medications were used less commonly. Intravenous fluid therapy was administered to 60 cases in total (60/371, 16.2%), with only 5 of these receiving an IV fluid bolus for the treatment of hypovolemic shock, and the remainder receiving a continuous infusion providing greater than or equal to maintenance fluid requirements. One dog with grade three clinical signs of tetrodotoxicosis received an IV bolus and 1 h infusion of IVLE, in addition to gastrointestinal decontamination.

Table 5.

Summary of dogs that presented multiple times for pufferfish exposure and/or presumed tetrodotoxicosis over the study period

Case ID	Breed, sex	Date	Number of pufferfish ingested	Grade of clinical signs	Inpatient vs. outpatient
169948	Labrador retriever MN	29.09.2014 28.09.2015	1 1	0 0	Outpatient Outpatient
148243	Labrador retriever MN	26.05.2012 21.12.2013	1 1	0 1	Outpatient Outpatient
118091	Labrador retriever FS	10.05.2009 04.01.2011	3 1	1 0	Outpatient Outpatient
161571	Labrador retriever FI	20.10.2013 08.01.2017	1 1	0 0	Outpatient Outpatient
97401	Staffordshire bull terrier FS	23.02.2007 19.01.2008 04.12.2008	1 1 1	0 0 0	Outpatient Outpatient Outpatient
137108	Labrador retriever FS	14.02.2012 14.04.2011	1 1	0 0	Outpatient Outpatient
162873	Border collie FS	18.09.2014 20.04.2015 10.04.2017	1 1 In mouth	1 0 1	Outpatient Outpatient Outpatient
87436	Siberian husky MN	19.12.2007 24.05.2008	1 3	3 4	Inpatient Inpatient
173306	Mixed breed MN	08.02.2015 26.03.2018	1 ?	0 3	Outpatient Inpatient
190644	Wheaten terrier MN	02.01.2017 11.12.2017	3 1	4 0	Inpatient Outpatient

Grade 0 is used to denote dogs that were asymptomatic. Grade 1 to 4 clinical signs are described in Table 1.

Eleven dogs were intubated in total. Eight received mechanical ventilation, two received manual ventilation prior to euthanasia (discussed in more detail in the outcome section) and one was intubated for management of upper respiratory tract obstruction (described above) without requiring ventilation. Hypercapnia was rarely documented as the specific indication for mechanical ventilation; rather, dogs were intubated and ventilated due to profound respiratory distress, generally before a blood gas analysis could be performed. Each of the patients requiring mechanical ventilation presented in or developed extremis. One medical record commented that ‘[the patient] presented in a panicked state, cyanotic, struggling to breathe but with minimal ventilatory effort…Intubated easily without drugs’. The median duration of ventilation was 10 h (Min‐Max 1–43). Complications observed in the cases requiring mechanical ventilation included aspiration pneumonia requiring antimicrobials in four dogs and central diabetes insipidus (CDI) in two dogs.

One of the dogs that developed CDI is described in the outcome section below because he also underwent cardiopulmonary resuscitation (CPR). The other dog diagnosed with CDI was a 10 –year‐old, male neutered Wheaten terrier, whose sodium concentration increased from 138 mmol/L (reference interval 140–150 mmol/L) at initial presentation to the referring veterinarian to 159 mmol/L 12 h later, during which time he was hyposthenuric (USG 1.003), and polyuric (urine output 11 mL/kg/h). A single dose of desmopressin (1ug/kg SC) and replacement of a 2 L free water deficit with 5% dextrose in water, resulted in normalisation of his sodium concentration (145 mmol/L), a decrease in urine output (5 mL/kg/h), and an increase in USG (1.018). No further treatment for CDI was required.

Multiple presentations

Ten dogs presented more than once during the study period for pufferfish exposure, with eight dogs presenting twice, and two dogs presenting three times. Of these dogs, four were asymptomatic at each presentation, whereas the other six had clinical signs of tetrodotoxicosis at one or more presentations. Table 4 summarises these cases, including the dates of presentation, number of pufferfish ingested, grade of clinical signs and inpatient versus outpatient treatment.

Table 4.

Counts and proportions of dogs receiving different treatments, classified by inpatient versus outpatient treatment after pufferfish ingestion and/or tetrodotoxicosis

	Summary statistics
Treatment	All dogs (n = 371)	Inpatient (n = 77)	Outpatient (n = 293)
No medications given	75, 20.2%	3, 3.9.6%	71, 24.2%
Antiemetics
Metoclopramide	196, 52.8%	27, 35.1%	169, 57.7%
Maropitant	31, 8.4%	17, 22.1%	14, 4.8%
Ondansetron	8, 2.2%	5, 6.5%	3, 1.0%
IV fluids
Anything other than a bolus	56, 15.1%	56, 72.7%	0
IV fluid bolus for shock	14, 3.8%	13, 16.9%	1, 0.3%
Antacids and gastroprotectants
Proton pump inhibitor	9, 2.4%	8, 10.4%	1, 0.3%
Ranitidine	5, 1.3%	2, 2.6%	3, 1%
Sucralfate	1, 0.3%	1, 1.3%	0
Intubation, ventilation and oxygen supplementation
Mechanical ventilation	8, 2.2%	8, 10.4%	0
Manual ventilation only	2, 0.5%	2, 2.6%	0
Nasal oxygen	2, 0.5%	2, 2.6%	0
Intubation without ventilation	1, 0.3%	1, 1.3%	0
Other	9, 2.4%	5, 6.6%	4, 1.4%

The total population displayed is 371 dogs since one dog was deceased on arrival and cardiopulmonary resuscitation was not performed.

Outcome

All but four dogs survived (369/372). One of the survivors had successful CPR in hospital and was neurologically intact at the time of discharge. Two dogs died naturally, both outside of the veterinary hospital, and two were euthanised. More information about the nonsurvivors is provided.

An 8‐year‐old, male neutered mixed‐breed dog (body weight 35 kg) vomited a single pufferfish 10 min after returning home from a walk by the river in Perth, WA, at 5 pm. Thereafter he continued to vomit, and the owner administered 2 diazepam tablets (dose unknown) at home. The dog then became unsteady in his hind legs, but was still able to wag his tail, so the owner went to bed. At 2 am, the owner awoke and found the dog collapsed outside, and so brought him to the clinic. The owner had observed the dog breathing when putting him into the car, but he was in cardiopulmonary arrest on arrival at presentation at TAHMU. Cardiopulmonary resuscitation was immediately commenced with routine basic life support and advanced life support, including IV adrenaline every 5 min for a non‐shockable rhythm, two doses of atropine and 1 L of IV lactated ringers' solution as a fluid bolus. Return of spontaneous circulation (ROSC) occurred after approximately 20 min. Post‐ROSC additional doses of adrenaline and atropine, as well as IV fluid boluses, were required to restore cardiovascular stability. Manual ventilation was transitioned to mechanical ventilation to maintain normoxia and normocapnia. Gastric lavage was performed but was not productive. Spontaneous ventilation attempts were first noted 12 h post‐ROSC. Concurrently he started to exhibit some response to stimuli (small leg and head movements), although cranial nerve responses remained absent. This prompted the addition of a midazolam continuous rate infusion, and later fentanyl CRI, for sedation. Moderate hypernatraemia (Na 167 mmol/L; reference interval 140–150 mmol/L) with marked polyuria and hyposthenuria (USG 1.003) developed consistent with CDI. Hypernatraemia resolved (Na 150 mmol/L), as did the polyuria (10 mL/kg/h to 2.7 mL/kg/h), with administration of subconjunctival desmopressin (~ 1ug/kg) and intravenous free water (5% dextrose in water), although the dog left the hospital less than 12 h after desmopressin dosing, and thus the need for ongoing treatment of CDI could not be determined. Twenty‐four hours post‐ROSC the dog developed hypotension, leukopenia and hypoglycaemia, which prompted the administration of a dopamine CRI, glucose supplementation and broad‐spectrum antimicrobials due to a concern for sepsis. He remained intubated without anaesthesia. This patient left the hospital against medical advice and was transported while still intubated by a veterinarian family member, but died in the car.

The other dog that experienced cardiopulmonary arrest before arrival at the veterinary clinic was a 5‐year‐old, spayed female mixed‐breed dog (20 kg). She had eaten 2 pufferfish while on a walk with her owners. They reported that she walked back to their car, vomited and then started panting. She died within 5–10 min while the owners were driving to AES. She was deceased on arrival and CPR was not attempted.

Two cases were euthanised. One was a 2‐year 7‐month‐old, intact female mixed‐breed dog (est. 40 kg) that presented to TAHMU with grade four clinical signs of tetrodotoxicosis 1–2 h after pufferfish ingestion at the beach. She was intubated and manually ventilated approximately 1 h after presentation due to the acute onset of cyanosis, but owners elected for euthanasia rather than ongoing treatment, due to a combination of financial concerns and the decision not to proceed with mechanical ventilation. The other was a 14‐year‐old, spayed female mixed‐breed dog (11.5 kg) that had a pufferfish in her mouth on a walk; she dropped the fish but shortly thereafter developed vomiting and lethargy. She was dyspnoeic on arrival and rapidly developed respiratory arrest, which prompted intubation and manual ventilation. Given her age, a recent poor recovery from an unrelated anaesthetic procedure, and financial limitations, the owner elected euthanasia rather than to pursue mechanical ventilation and ICU care.

Discussion

The findings of this retrospective case series are supportive of our first hypothesis that clinical signs of presumed tetrodotoxicosis in dogs are generally mild with a good prognosis for recovery, with just over two thirds of symptomatic dogs having grade one signs only, as well as a low rate of hospitalisation, short duration of hospitalisation and high survival rate. The findings were also supportive of our second hypothesis that the ingestion of multiple pufferfish and a delay to presentation at a veterinary hospital are associated with a greater likelihood of developing clinical signs. However, some dogs developed severe clinical signs very rapidly after exposure to a single pufferfish, and so this did not always hold true.

Despite the study hospitals being emergency and referral centres, the majority of dogs in our study were primary accession cases, suggesting that our findings are likely to be representative of what is seen in practice. Only a small number of cases were referred; some of which had life‐saving interventions at their primary veterinarian, including intubation and ventilation. The ability of primary care veterinarians to facilitate the transport of intubated and ventilated dogs for ongoing mechanical ventilation undoubtedly saved the lives of the four dogs in this scenario in our study. These cases, along with the association between the time to presentation and development of tetrodotoxicosis, reinforce the importance of dogs receiving early veterinary care after pufferfish ingestion. It is likely that many pet owners in Australia are aware of the risk pufferfish ingestion poses to dogs, since so many of them came straight from the beach or river to a veterinary hospital after ingestion.

The dogs that ingested pufferfish in this study were typically young dogs of common breeds known to be both food motivated and likely to scavenge, such as Labrador retrievers and Staffordshire bull terriers. Other case series of toxicoses have described similar breed profiles. ²⁷ , ²⁸ Nonetheless, a wide variety of breeds were represented. Body weight suggested that medium‐to‐large dogs were affected, with no clinically relevant difference between those remaining asymptomatic and those developing clinical signs of tetrodotoxicosis, although this was not statistically compared. The location of pufferfish ingestion was the beach in 40–50% of cases, with approximately 20% having exposure to pufferfish from inland waterways, and the remainder not reported. Although not specifically determined, it is suspected based on the location of the study hospitals that the inland waterways (e.g. rivers, canals) where dogs were exposed to pufferfish were saltwater or brackish water, rather than fresh water. Indeed, TTX is best characterised as a toxin found in marine organisms including marine pufferfish, but has been described in freshwater fish species as well, including freshwater pufferfish. ²

More than 80% of dogs ingested a single pufferfish, but the ingestion of multiple pufferfish did occur. The occurrence of clinical signs of presumed tetrodotoxicosis despite some dogs only licking, chewing or holding a pufferfish in their mouths is consistent with the known presence of TTX in pufferfish skin. ² In fact, even one of the dogs that was euthanised due to the requirement for mechanical ventilation only had a pufferfish in her mouth and did not ingest a fish. Factors contributing to toxicosis in individual dogs likely include a combination of the amount of TTX in the ingested material and as well as patient factors. For example, the aforementioned dog that was euthanised due to requirement for mechanical ventilation after just having a pufferfish in their mouth was small and elderly. On the other hand, the two dogs that ate 12 pufferfish and only developed grade one clinical signs were young Labradors.

When collecting data for our study, we considered the question of how tetrodotoxicosis should be defined in veterinary patients and ultimately concluded that the similarities with the toxidrome in humans allowed application of a similar grading system. In human medicine, where pufferfish is eaten as part of a prepared meal, gastrointestinal upset is attributed to the effects of the systemically absorbed TTX on the chemoreceptor trigger zone (CRTZ). ²⁹ Experimental studies in dogs ¹⁵ and cats ³⁰ have also shown that TTX induces vomiting via its effects on the CRTZ. Direct effects of TTX by binding to voltage‐gated sodium cannels in the enteric nervous system may also contribute to gastrointestinal signs. ³¹ Another contributor to gastrointestinal upset in some cases could be the putrid nature of the ingested fish that had potentially been caught and discarded by fisherman hours or even days earlier. Additionally, the pufferfish may have been ingested along with other materials, such as sand, saltwater or seaweed. The contribution of these factors to direct gastric stimulation of vomiting is unknown. Nonetheless, given that the vomiting of other foreign material was not usually evident in these cases, we thought it appropriate to classify dogs with gastrointestinal signs as having grade one tetrodotoxicosis. It is likely that some dogs classified as asymptomatic actually had sensory signs of tetrodotoxicosis, consistent with grade one or even grade two tetrodotoxicosis in humans, that went undetected. Antiemetic therapy was given to the majority of dogs in this case series since many dogs continued to vomit even after producing the pufferfish that they had ingested.

Signs of ataxia and weakness progressing to tetraparesis, and requiring ventilation, seem to be similar in dogs with tetrodotoxicosis to those reported in humans. Consistent with the mechanical ventilation literature in dogs, ³² increased work of breathing was the most common reason for the commencement of mechanical ventilation in our cases, without necessarily documenting hypoxemia or hypercapnia prior. Fortunately, only approximately 2 per 100 pufferfish exposed dogs required mechanical ventilation. Additionally, the duration of ventilation needed was short and the prognosis excellent compared with other causes of lower motor neuron signs that require ventilation such as tick paralysis. ²³ , ³³ , ³⁴ Another nervous system sign observed in dogs in our case series that has also been described in humans is that of CDI. Central diabetes insipidus was diagnosed in two dogs in our case series, although one was likely secondary to brain injury after protracted cardiopulmonary arrest, rather than tetrodotoxicosis. The other case however had no other reason for CDI than tetrodotoxicosis, similar to a human case report. ³⁵ The authors of that report suggested that blockade of sodium channels of the magnocellular neurons in the neurohypophysis may have resulted in transient inhibition of arginine vasopressin release leading to CDI. ³⁵

Pufferfish poisoning has been previously reported in about 30 people in Australia. ⁶ Case series and case reports have also described human cases of tetrodotoxicosis from Japan, China, Taiwan, Thailand, Bangladesh and the United States. ² , ⁸ , ³⁶ The largest case series in Australia included 11 cases, of which nine were from a single family, identified via a search of the New South Wales Poisons Information Centre. ⁶ Clinical signs in these patients ranged from none (one child was completely asymptomatic), to mild hyperesthesia (perioral and extremity), to respiratory failure requiring intubation and ventilation in one patient. Several patients also had vomiting, but none had cardiovascular symptoms. Another publication has documented pufferfish ingestion and tetrodotoxicosis in three people from the same family in Western Australia. ⁷ Not only were the three humans symptomatic, but they also fed pufferfish to their dog who vomited immediately. ⁷ All three of the humans developed neurologic signs commencing within 1 h of ingestion; in two cases, these signs were so severe that the patients could not stand by the time that they presented to the emergency room 6 h later. Interestingly only one of the affected humans had vomiting as a clinical sign, so this report may suggest that dogs vomit more readily than humans with TTX exposure and rapid expulsion of the TTX containing fish (self‐decontamination) may mitigate the severity of subsequent clinical signs in dogs. The proportion of all exposed dogs in our study that had vomiting (40.1%) was nonetheless similar to that in a large case series in Bangladesh where 29/63 (46%) of affected humans had nausea/vomiting. ³⁶

The only unusual clinical sign noted in dogs that ingested pufferfish that has not been clearly described in humans was facial angioedema in two dogs. Whether or not this was related to a type one hypersensitivity reaction to a component of the ingested pufferfish, the TTX itself or something else that these dogs were exposed to is unknown. Regardless, both dogs ultimately had a good outcome with supportive treatment for their tetrodotoxicosis and presumed anaphylaxis.

The onset of clinical signs in our case series also appears similar to that described in humans. Clinical signs in humans develop rapidly, with the onset of grade one signs within 5–45 min, ¹² but early effects are usually sensory, which could be missed in dogs. Onset of severe signs in humans has been described as 15 min to several hours (grade 3), and 15 mins to 24 h (grade 4) with duration of signs lasting from a few hours (minor cases) to days (moderate to severe cases). ¹² The time course of recovery also appears similar in dogs, with milder cases recovering within hours. Even severely affected dogs requiring mechanical ventilation had a short duration of hospitalisation of a couple of days. Early gastrointestinal decontamination in the majority of dogs is likely to have prevented the development of clinical signs in some asymptomatic dogs and prevented progression to grade three and four clinical signs in symptomatic dogs.

To the knowledge of the authors the only other veterinary case series of tetrodotoxicosis is from New Zealand. ¹⁸ McNabb and colleagues described an outbreak of neurotoxicosis in dogs near Auckland, between July and September 2009. Fifteen affected dogs were described with clinical signs including vomiting (14/15, 93%), salivation (13/15, 87%), ataxia (12/15, 80%), lethargy/sedation (11/15, 73%), cardiac arrhythmias (8/15, 53%), diarrhoea (6/15, 40%), respiratory failure (4/15, 27%) and muscle fasciculations (3/15, 20%). Five of the affected dogs (33%) died. While this study describes an exhaustive epidemiologic and toxicologic investigation, no information about the veterinary care provided to the affected dogs, or the time course of their illness was reported. Tetrodotoxin from the grey side‐gilled sea slug (Pleurobranchaea maculata) was ultimately implicated as the aetiologic agent of the toxidrome, and TTX was identified in dog vomitus and stomach contents. In total, the investigators screened for 26 marine biotoxins. It is clear that the dogs in this New Zealand report had a more severe toxidrome than the dogs included in our Australian case series, however the pattern of clinical signs was very similar in those that were affected. The difference in severity could be due reporting bias in the New Zealand paper (i.e. Only the most severely affected dogs recognised as having the toxidrome), a greater concentration of TTX in the grey side‐gilled sea slug than in pufferfish, or due to ingestion of a large number of sea slugs by each dog, given the description of these organisms washing up on the beaches in large numbers at the time of the outbreak. The New Zealand case series is a pertinent reminder that although pufferfish are classically considered the culprit in tetrodotoxicosis, numerous other marine organisms can contain the toxin and hence pose a threat to dogs if ingested.

Activated charcoal is known to adsorb TTX in vitro, ³⁷ and ex vivo from human body fluids, ³⁸ and hence the use of activated as part of gastrointestinal decontamination in many dogs in this case series is considered appropriate. It was unclear why a portion of dogs were not treated with activated charcoal, but it may be that veterinarians assumed that vomiting up the ingested pufferfish was adequate. Indeed, no studies have evaluated the potential benefit of activated charcoal after the ingested pufferfish has been expelled or the time window during which activated charcoal may be effective. Nonetheless, given that activated charcoal is generally considered safe and many dogs will voluntarily eat it, the authors recommend use of activated charcoal in cases of pufferfish ingestion, after emesis, where no contraindications exist. The main adverse effect of activated charcoal with sorbitol in some studies is hemoconcentration and hypernatraemia, ³⁹ which may be why some dogs in this case series were administered intravenous fluids during hospitalisation. Pneumonia has also been reported as a complication of aspirating activated charcoal in a dog, ⁴⁰ and thus the potential benefit versus risk of its use in the management of dogs after pufferfish ingestion should be weighed on a case by case basis.

In contrast to activated charcoal, there is no evidence that IVLE is a useful tool for the decontamination of dogs with tetrodotoxicosis. The purported mechanisms of action of IVLE include acting as a lipid sink for lipophilic toxins, serving as an energy source facilitating mitochondrial recovery in local anaesthetic toxicity and increasing intracellular calcium in cardiomyoctes for calcium channel blocker toxicity. ⁴¹ , ⁴² None of these mechanisms is likely to be beneficial in dogs with tetrodotoxicosis. Tetrodotoxin is strongly hydrophilic with a predicted logP of −5.9. ⁴³

Limitations of this study are due to its design as a retrospective study and reliance on completeness and accuracy of medical records. Toxicologic testing was not performed in contrast to many studies in humans and the aforementioned study in dogs in New Zealand, and follow‐up was also limited. It is possible that some dogs with tetrodotoxicosis were missed due to the inclusion criteria requiring the owner or veterinarian to have witnessed a pufferfish. For example, some cases with a presumptive diagnosis of tetrodotoxicosis, including some that were mechanically ventilated, did not meet inclusion criteria despite possible exposure (e.g. being near waterway) and having an acute onset of paralysis. This included cases in Queensland with a negative tick clip, which refers to systematic clipping of the fur coat to facilitate a thorough visual and tactile examination of the skin for attached ticks, to rule out paralysis due to Ixodes holocyclus.

Conclusion

Consistent with our hypothesis, clinical signs are generally mild when presumed tetrodotoxicosis occurs after pufferfish ingestion in dogs. The prognosis for complete recovery is excellent with timely initiation of gastrointestinal decontamination and supportive care.

Conflicts of interest and sources of funding

The authors have no conflicts of interest relevant to this manuscript.

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.