History

A 5-year-old 26.7-kg castrated male Basset Hound was referred for evaluation of hyporexia and icterus. The dog had a prolonged history of dietary indiscretion, weight loss, diarrhea, and inflammatory bowel disease. The inflammatory bowel disease was being managed with prednisone (1 mg/kg, PO, q 12 h) and cyclosporine (5.6 mg/kg, PO, q 12 h). Two weeks prior to presentation the dog exhibited signs of neck pain and was diagnosed with presumptive intervertebral disc disease and started on gabapentin (10 mg/kg, PO, q 8 to 12 hours). On initial physical examination, the dog was icteric, had a grade 2/6 left apical systolic murmur, and was quiet, alert, and responsive. The remaining findings on physical examination were unremarkable. The dog was admitted to the hospital for supportive care and additional diagnostic workup.

Initial hematology revealed marked increases in liver enzyme activities (alanine aminotransferase, 3,139 U/L; reference range, 16 to 91 U/L; aspartate aminotransferase, 730 U/L; reference range, 23 to 65 U/L; alkaline phosphatase, 1,125 U/L; reference range, 20 to 155 U/L; and γ-glutamyltransferase, 287 U/L; reference range, 7.0 to 24.0 U/L), hyperbilirubinemia (5.7 mg/dL; reference range, 0.1 to 0.5 mg/dL), and hypercholesterolemia (512 mg/dL; reference range, 128 to 317.0 mg/dL). A CBC revealed a marked leukocytosis (23.57 X 10³ WBCs/µL; reference range, 5.7 X 10³ to 14.20 X 10³ WBCs/µL) characterized by a neutrophilia (20.03 X 10³ cells/µL; reference range, 2.7 X 10³ to 9.4 X 10³ cells/µL) with a left shift (band neutrophils, 0.47 X 10³ cells/µL; reference range, 0.0 X 10³ to 0.1 X 10³ cells/µL). A coagulation profile revealed prothrombin time and partial thromboplastin time within reference limits but a slightly high concentration of D-dimers (0.46 µg/mL; reference range, 0.0 to 0.2 µg/mL).

Abdominal ultrasonography revealed an enlarged, diffusely hyperechoic liver. The gallbladder was of normal size but had a thickened wall and a diffuse transmural hyperechoic appearance. The common bile duct and intra- and extrahepatic ducts were of normal diameter. Although some of the appearance may have been due to steroid hepatopathy, the appearance was mostly consistent with cholangitis or cholangiohepatitis. On thoracic radiography, cardiovascular structures were within normal limits; however, a mild bronchovesicular pattern and enlarged liver were noted.

The patient was anesthetized for laparoscopic liver biopsies and cholecystocentesis. Prior to induction of anesthesia, a 3-lead ECG, pulse oximeter, and oscillometric blood pressure device were instrumented for monitoring during induction and maintenance of anesthesia. Anesthesia was induced with fentanyl (5 µg/kg, IV) and alfaxalone (2 mg/kg, IV), and constant rate infusions of remifentanil (0.2 to 0.3 µg/kg/min, IV), lidocaine (50 µg/kg/min, IV), and alfaxalone (100 µg/kg/min, IV) were started. Crystalloid IV fluid therapy was initiated with lactated Ringer solution (2 mL/kg/h). An endotracheal tube was premeasured to the level of the thoracic inlet to ensure proper depth of placement. Following endotracheal intubation, a capnograph device was applied to the end of the endotracheal tube and a waveform was identified. The cuff was inflated by use of a positive pressure leak check to a maximum pressure of 15 cm H₂O. The free end of the endotracheal tube was connected to a rebreathing circle system delivering 100% O₂ (2 L/min), and the patient was allowed to breathe spontaneously. Due to the conformation of the patient, peripheral IV access was challenging to maintain; therefore, a triple-lumen central IV line was placed in the right jugular vein. An arterial catheter was placed in the right dorsal pedal artery to facilitate direct blood pressure monitoring. Once the patient was moved into the operating theater and positioned in dorsal recumbency, the alfaxalone and lidocaine infusions were discontinued, and anesthesia was maintained with desflurane (end-tidal desflurane concentration, 2.5% to 7.2%) delivered in O₂ to effect and remifentanil infusion (0.2 to 0.3 µg/kg/min, IV). Mechanical ventilation was initiated with a tidal volume of 10 mL/kg, respiratory rate of 10 breaths/min, and peak inspiratory pressure (PIP) of 13 cm H₂O. The patient was intermittently hypotensive (mean arterial blood pressure < 60 mm Hg) throughout the anesthetic episode. This was managed with the use of 2 crystalloid fluid boluses (6 mL/kg total), a dopamine constant rate infusion that was titrated to effect (7 to 15 µg/kg/min, IV), and atropine (0.01 mg/kg, IV). During this time, the patient had clinically normal end-tidal CO₂ concentration (ETco₂; 41 to 50 mm Hg) and O₂ saturation of hemoglobin (Spo₂; 94% to 98%), as measured by pulse oximetry.

Thirty minutes following insufflation of the abdomen, placement of laparoscopic ports, and laparoscopic collection of the liver biopsies, while the cholecystocentesis was being performed, the PIP rose to 15 cm H₂O and patient-ventilator dyssynchrony was noted, with the patient becoming progressively tachypneic. Mechanical ventilation was discontinued, and the patient rapidly desaturated (Spo₂ of 72%), developed relative tachycardia (heart rate increased from 100 to 140 beats/min), and became hypotensive (mean arterial pressure [MAP] fell from 90 to 60 mm Hg) while the ETco₂ remained stable at 49 mm Hg.

Question

What could have accounted for the rapid development of hypoxemia, tachypnea, tachycardia, and hypotension?

Answer

A tension pneumothorax was the primary differential diagnosis for the tachypnea, hypoxemia, tachycardia, and hypotension in the face of a normal ETco₂ seen in this patient. Hypoxemia secondary to the development of tension pneumothorax is multifactorial. The accumulation of gas in the pleural space between the lungs and chest wall leads to collapse of the lung, causing an intrapulmonary shunt, which accounts for the desaturation despite administration of 100% O₂ with desflurane for an end-tidal desflurane concentration of 2.5% to 7.2%. As pressure built within the thorax, there was also a decline in venous return, which reduced stroke volume and cardiac output leading to a decline in the MAP.¹ The ETco₂ was 50 mm Hg before the development of the pneumothorax, then declined to 27 mm Hg during the pneumothorax, and then climbed to 37 mm Hg following resolution; this was likely due to dead space ventilation related to alterations in cardiac output secondary to decreased venous return. Tachycardia was likely a combined effect secondary to both hypoxemia and hypotension.²

Thoracic auscultation revealed a lack of bronchovesicular sounds over the left hemithorax; thus, the laparoscopic procedure was aborted, laparoscopic ports were removed, and abdomen was opened using a ventral midline incision releasing the abdominal insufflation. A 2-cm incision was made in the ventral muscular portion of the diaphragm and opened with hemostats. Immediately, the Spo₂ improved to 98%, MAP improved to 70 mm Hg, and PIP decreased to 10 cm H₂O. A chest tube was placed in the left hemithorax, and following closure of the diaphragm, the chest tube was aspirated until negative pressure was obtained. Intermittent positive pressure ventilation was provided during the management of pneumothorax until resolved; afterward, the patient breathed spontaneously for the duration of the procedure, during which time the patient’s vital signs remained stable. The patient was extubated 5 minutes after the completion of the surgical procedure. Methadone (0.2 mg/kg, IV) was administered for postoperative analgesia. Arterial blood gas analyses during the event and in recovery would have been desirable to assess the degree of hypoxemia; however, a sample was not collected during management and correction of the pneumothorax due to limited personnel. Immediately following recovery, the arterial line functionality was lost during transport of the patient.

Following recovery from anesthesia, the patient was transferred to the intensive care unit for further monitoring and care. Subsequent aspirations of the chest tube yielded small volumes of serosanguinous fluid and less than 2 mL of air; thus, the chest tube was removed 24 hours after placement. The dog was started on broad-spectrum antimicrobials and was ultimately discharged from the hospital 6 days after surgery.

Discussion

In humans, the development of tension pneumothorax during laparoscopic cholecystectomy has been described,^3–7 with tension pneumothorax in affected patients attributed to inadvertent insufflation of the thorax via the abdomen with CO₂; the condition is termed “carbothorax.” When the gas in the thorax was sampled from these cases, it was noted to have been CO₂, not O₂, suggesting that the source of the gas was CO₂ within the abdomen that gained access to the thorax through a defect in the diaphragm. Because the abdomen is maintained under constant pressure, CO₂ is continuously insufflated into the abdomen to compensate for the gas lost into the thorax, thereby causing a tension pneumothorax. A model of this has been studied in pigs.⁷ Animals were anesthetized and underwent peritoneal insufflation with CO₂, and a laceration was created in the left diaphragm; researchers found that there were substantial declines in arterial partial pressure of O₂ and Spo₂ and increases in arterial partial pressure of CO₂ and PIP.⁷

In humans, the appearance of a “floppy diaphragm” is an indicator of the development of a pneumothorax during laparoscopic procedures. The diaphragm billows inferiorly due to the loss of negative pressure within the thorax as the abdominal insufflation is released, thus denoting the resolution of the “tension” portion of the tension pneumothorax.⁸ The development of carbothorax during laparoscopic procedures is usually attributed to preexisting defects of the diaphragm, iatrogenic trauma to the diaphragm, and diffusion of CO₂ through anatomic pathways.⁸ For the dog of the present report, once the abdomen was opened and insufflation lost, the surgeons reported that the diaphragm had a blue-gray appearance and was described as billowing inferiorly, as reported in the human case reports.

In our patient, we suspected that the needle used for cholecystocentesis was inadvertently advanced through the gallbladder, liver, and diaphragm, causing a pathway for CO₂ to travel from the abdomen into the thorax. This coincided with the rapid development of the pneumothorax during this portion of the procedure. Other possible differential diagnoses for the observed clinical signs included pulmonary thromboembolism and non-CO₂ pneumothorax. Although we did not analyze the gas collected upon release of the pneumothorax, the idea that this was a CO₂ pneumothorax secondary to abdominal insufflation was supported in multiple ways. First, thoracic radiography prior to anesthesia did not reveal important pulmonary parenchymal disease that would have predisposed this patient to development of pneumothorax due to mechanical ventilation. Second, no gas was aspirated from the chest tube following evacuation of the chest at surgery; this would have indicated that air leakage from another source was unlikely. Additionally, this patient was never noted to be hypoxemic again, as evidenced by normal pulse oximetry readings for the duration of care. Although pulmonary thromboembolism can be a catastrophic and important cause of hypoxemia in anesthetized patients, it is initially characterized by a sudden and dramatic decline in ETco₂ secondary to an increase in dead space ventilation. In this patient, the decline in ETco₂ was the last clinical sign appreciated, moving it further down the list of potential differential diagnoses.

Although there are numerous benefits associated with performing laparoscopic procedures, they also present several anesthetic challenges due to alterations in hemodynamic and respiratory function associated with abdominal insufflation. The development of a carbothorax should be considered as a potential cause of any sudden decrease in Spo₂ with a simultaneous decline in arterial blood pressure and increase in heart rate in animals anesthetized for laparoscopic procedures in which insufflation of the abdomen is performed using CO₂.