Fatal Venous Air Embolism During Anesthesia in an Apparently Healthy Adult Chihuahua*

Introduction

Intraoperative venous air embolism has been extensively described in the human medical literature in a wide variety of clinical scenarios. It is less commonly described in dogs and cats and has been seen during pneumocystography, laparoscopy, pharyngoscopy, oral surgery involving either pressurized air drills or cryosurgery, and inadvertent air administration through an IV catheter.^1–8 Clinical signs and cardiovascular arrest occur rapidly and appear irreversible based on the fact that all published veterinary cases died despite attempts at cardiopulmonary resuscitation. In suspected cases of venous air embolism, air can be demonstrated within the cardiovascular system either via radiography or by careful assessment for IV/intracardiac air bubbles during postmortem examination. This case report describes inadvertent, iatrogenic air administration through an IV catheter during anesthesia leading to fatal venous air embolism in an apparently healthy toy-breed dog. The necropsy technique used to detect air in the right ventricle is demonstrated.

Case Report

A 5 yr old female Chihuahua weighing 2 kg was presented for elective ovariohysterectomy at the owner's request. The nulliparous bitch had no known previous health problems, had a body condition score of 5 out of 9, and was current on vaccinations. The only significant finding on physical examination was mild to moderate dental tartar and gingivitis, predominantly associated with retained deciduous upper canine teeth. The right upper deciduous canine tooth was loose. Preanesthetic blood work was recommended but declined by the owner.

On the morning of surgery, the fasted bitch was premedicated with dexmedetomidine^a (5 μg/kg) and morphine^b (0.5 mg/kg) intramuscularly. A standard 22-gauge IV catheter was placed in a cephalic vein with a T-connector^c attached and anesthesia was induced with propofol^d (4.5 mg/kg) and maintained with isoflurane^e delivered in O₂. A balanced electrolyte solution^f was administered IV at a rate of 5 mL/kg/hr starting in the surgical preparation area. Doppler blood pressure^g, pulse oximetry^h, end-tidal CO₂, rectal temperature, heart rate, and respiratory rate were monitored. The patient was shaved and prepared for abdominal surgery, disconnected from IV fluids, then transported to the operating room. Doppler blood pressure (100–120 mm Hg), temperature (100–100.3°F), heart rate (80–85 beats/min), respiratory rate (8–20 spontaneous breaths/min), end-tidal CO₂ (30–35 mm Hg), and pulse oximetry (100%) were considered appropriate (parenthetic values are recorded ranges for the patient throughout this period).

An IV fluid setup consisting of a 500 mL bag of a balanced electrolyte solution, an infusion setⁱ, and a fluid infusion pump^j was connected directly to the T-connector on the patient's IV catheter on arrival in the operating room. As soon as the fluid pump was activated, it began an audible alarm while flashing a message that there was air in the infusion line. Mistakenly thinking that the alarm was due to an occlusion, the anesthesia technician manually flushed the IV set with a previously prepared 3 mL syringe of sterile saline at an injection port one-quarter of the way up the infusion line from the patient. The patient abruptly convulsed and became apneic with cyanotic mucous membranes. Blood pressure was undetectable and the electrocardiogram showed a sinus bradycardia. Atipamezole^k (0.02 mL) was administered, one-half of the dose IV and the remainder intramuscularly, as was naloxone^l (dose not specified). Manual positive-pressure ventilation with 100% O₂ was provided. The patient remained bradycardic with poor femoral pulses and progressed rapidly to full cardiac arrest. Cardiopulmonary resuscitation (CPR) was initiated, including alternating thoracic and abdominal compressions. Two IV boluses of atropine^m (0.04 mg/kg) and three IV boluses of epinephrineⁿ (two boluses at 0.025 mg/kg and one bolus at 0.013 mg/kg) were administered. External defibrillation was attempted. Resuscitation was unsuccessful, and the patient was pronounced dead after 20 min of CPR. The body was submitted for necropsy examination.

The necropsy examination, performed 3–4 hr after death, identified soft-tissue hemorrhage in the right ventrolateral cervical region associated with the right external jugular vein. Blood released from the incised vein was frothy. The pericardium was incised to create a pericardial window, the pericardial sac was filled with water to submerge the heart, and the right ventricular free wall was punctured. Several air bubbles emerged from the right ventricle through the puncture hole (Supplementary Video I). Additional gross findings included moderate dental calculus; a retained upper left deciduous canine tooth; mild, smooth, nodular thickening along the free margin of the left atrioventricular valve leaflets (endocardiosis); mild anthracosis; focal liver fractures with hemoperitoneum; and dorsoventral tracheal narrowing. The loose upper right deciduous canine tooth identified clinically had presumably been lost prior to necropsy. In addition to microscopic features of the above gross findings, histological lesions included minimal hepatic lipidosis, rare renal tubular epithelial necrosis, and mild lymphoplasmacytic gastroenteritis. Based on the clinical history and the presence of air within the right ventricle of the heart at the time of necropsy, venous air embolism during anesthesia was considered the most likely cause of death.

Discussion

Although it is generally understood that IV administration of air can be fatal, only a few documented clinical reports exist in veterinary medicine. A 9 mo old Pomeranian that suffered a suspected iatrogenic venous air embolism through an IV catheter during anesthesia for fracture repair also had pulmonary calcification detected at the time of necropsy.⁸ Similar to the present case, air was demonstrated postmortem in the cardiovascular system by submerging the intact heart in water, incising the right ventricle, and observing the release of air bubbles from the ventricular chamber. The authors of that case report posited that the pulmonary mineralization thought to be related to renal dysplasia may have increased the pet's sensitivity to venous air embolism. A geriatric Weimaraner went into cardiopulmonary arrest during nitrogen gas insufflation of the peritoneal cavity for laparoscopy.⁴ At necropsy, the spleen contained a small puncture hole and had gas bubbles within the parenchyma on cut section. The Weimaraner also had emphysematous bulla of the lung and several adult heartworms in the right ventricle and pulmonary artery. The authors did not comment on the potential contributing role of the underlying cardiopulmonary pathology to mortality, perhaps because the large quantity of insufflated air was considered sufficient to be fatal independent of any pre-existing lesions. Two adult dogs, one a miniature poodle and the other a borzoi, arrested during cryosurgery for treatment of gingival neoplasms while pressurized liquid nitrogen was administered into the tumor bed using a spray technique.⁷ Air was demonstrated radiographically in the right atrium and right ventricle of each dog, supporting a diagnosis of fatal air embolism. No clinicopathological abnormalities or pre-existing conditions were described in those two dogs. Fatal air emboli have been associated with pneumocystography in several cases, including a 3.5 mo old Brittany spaniel and three adult cats.^1–3 In the Brittany spaniel puppy, air was detected radiographically in abdominal veins and a “churning” sound was ausculted in the heart shortly before cardiac arrest.¹ In two of the cats, thoracic radiographs taken after cardiopulmonary arrest confirmed air in the right ventricle of the heart.^2,3 In the third cat, air was detected in blood vessels surrounding the kidneys at the time of necropsy.³ In all four animals undergoing pneumocystography in the above-mentioned reports, pre-existing pathology was limited to the lower urinary tract. A young adult cat with a retropharyngeal diverticulum arrested following gas insufflation of the diverticulum during pharyngoscopy.⁵ Gas was detectable radiographically and ultrasonographically in the right atrium and ventricle, abdominal vena cava, and hepatic veins postmortem. The patient had no detectable pre-existing or underlying pathology apart from the diverticulum. An adult cat undergoing multiple dental extractions arrested and died following venous air embolism.⁶ The authors suggested that massive amounts of air were introduced into the vasculature through the use of an air-driven dental drill. Again, no pre-existing conditions other than stomatitis were described in this patient.

The Chihuahua in the present case had very mild pre-existing lesions in the heart and respiratory system detected on gross examination, including endocardiosis, anthracosis, and dorsoventral tracheal narrowing. Those lesions were considered to be either incidental or subclinical and unlikely to contribute to mortality. The liver fractures and hemoperitoneum were attributed to trauma from attempted cardiopulmonary resuscitation. Hemorrhage surrounding the right external jugular vein likely reflected venous blood draw attempts at that site during CPR. The presence of frothy blood within that vein may have been from IV injections during CPR or from regurgitation from the right heart during thoracic compression.

The severity of clinical signs of venous air embolism is related to the volume of gas delivered and the rate of entrainment of gas into the circulatory system.^9,10 Death from venous air embolism is thought to result from either pulmonary vascular constriction/hypertension with smaller volumes or mechanical air lock obstructing outflow from the chambers of the right heart with larger volumes.^9,10 Changes detected via anesthetic monitoring include early peaking of P waves followed by S-T segment depression, bradycardia or other arrhythmias, altered heart sounds, increased central venous pressure, decreased systemic arterial blood pressure, decreased end-tidal partial pressure of CO₂, decreased arterial O₂ saturation, and mucous membrane pallor or cyanosis.^{1,2,4–6,8,9,11} Air can be demonstrated within the heart and vasculature either radiographically or ultrasonographically before or shortly after death.^{1–3,5–7,10} Alternatively, intact vasculature can be carefully evaluated during necropsy examination for intraluminal gas bubbles or the intact heart can be submerged in water prior to incising the right chambers to assess for the release of air. Postmortem detection depends on an index of suspicion because IV air may go undetected using routine dissection techniques.

Interestingly, the majority of published case reports of fatal air embolism in dogs and cats describe air introduced into the cardiovascular system via indirect methods (gas insufflation of a body cavity or hollow viscous or the use of air-pressurized instruments in the oral cavity) rather than by direct injection during venipuncture or through an IV catheter. Presumably, small air bubbles are injected through IV catheters with some frequency during routine anesthesia and in-hospital medical care with no apparent adverse effects. In an experimental study of vascular air emboli in dogs, death occurred in animals receiving 0.69 mL of IV air/kg body weight/min but not in study groups receiving air infusions ranging from 0.13–0.61 mL/kg/min.¹¹ In a separate experimental study, dogs receiving between 1.2 and 1.8 mL/kg/min often survived if the infusion was short lived (<3 min), whereas dogs receiving >1.8 mL/kg/min typically did not recover.⁹ In dogs, the reported lethal volume of air injected as a single bolus ranges from 7.5 to 15 mL/kg.^12,13 In the present case and the previously reported case of suspected iatrogenic venous air embolism, both dogs were toy breeds weighing ≤2 kg.⁸ The volume of air administered to each dog in those two cases was not known; however, the dog in the present case may have received a maximum of 5–7 mL of air (assuming an air-filled line between the injection port and patient was completely flushed to the patient). It is possible that inadvertent injection of the same quantity of air in a larger breed would have been inconsequential.

Conclusion

This case describes inadvertent IV air administration causing fatal venous air embolism in an apparently healthy toy-breed dog undergoing anesthesia for an elective procedure. Air was confirmed in the cardiovascular system at the time of necropsy by submerging the intact heart in water prior to puncturing the right ventricle to release air bubbles. Only one published veterinary case of fatal venous air embolism was due to suspected direct IV air administration, suggesting that this etiology is either rare or underreported.⁸ In addition, the small size of the patient in that report and in the current case may represent a predisposing factor.⁸ The present case emphasizes the importance of adequately priming IV infusion sets and carefully attending to fluid pump alarms. Although very uncommon, fatal venous air embolism should be a differential diagnosis in animals dying from abrupt cardiovascular arrest either during anesthesia or following administration of IV injections, particularly in small patients. Postmortem thoracic radiography or specific necropsy techniques as described above should be employed in suspect cases of fatal venous air embolism to increase the probability of achieving a definitive diagnosis.

Video I – The opened thoracic cavity contains the heart in situ. The intact heart is submerged in water within the incised pericardial sac. The right ventricular free wall is punctured with a scalpel blade, allowing the release of several air bubbles from the right ventricular lumen.