EXTERNAL heating by conventional electric cautery or by laser provides coagulation of the tissue by conducting the heat from the surface to the inside of the tissue. Conversely, internal heating is the method for necrotizing tumor and hemostasis by means of a dielectric heating electrode, which is inserted into the tissue or tumor. The two methods of administering localized internal thermal therapy of the liver are microwave coagulation therapy (MCT) 1and radiofrequency ablation (RFA). 2,3
A venous air embolism, a potential complication of surgery in the sitting position 4or of laparoscopy, 5,6is unlikely to occur during laparotomy in a horizontal position. However, we experienced a case with abrupt decreases in end-tidal carbon dioxide partial pressure (Petco2) and hemoglobin oxygen saturation (Spo2) during laparotomic MCT in the left lateral recumbent position, probably due to pulmonary gas emboli. We believe that the case discussed below is the first detailed report of a pulmonary gas embolism caused by localized thermal therapy of the liver.
A 71-yr-old man (weight, 44 kg) was scheduled to undergo right segmentectomy of the liver for the removal of a single hepatoma nodule (2 × 3 cm) under general anesthesia. The patient, classified as American Society of Anesthesiologists (ASA) physical status 2 because of emaciation, had no contributory past or family history, and the preoperative laboratory data were within normal ranges. After premedication with atropine and hydroxyzine and placement of both standard ASA monitors and an intraarterial catheter, an epidural catheter was inserted at T8–T9. Anesthesia was induced with fentanyl, thiamylal, and vecuronium; thereafter, anesthesia was maintained with isoflurane in a 1:1 gas mixture of oxygen and nitrous oxide, and supplemental fentanyl and vecuronium in the left lateral decubitus position. Petco2was 25–30 mmHg.
After the right segmentectomy, MCT (Microtaze® Model OT-110 M, Nippon Shoji, Osaka, Japan) was repeatedly applied to the liver for hemostasis at the output of 60 W and the frequency of 2,450 MHz for 60–120 s at a time. After several applications of MCT, an abrupt decrease in Petco2from 27 to 10 mmHg occurred with concomitant decrease in Spo2from 100 to 96%. The blood pressure was stable, and the electrocardiogram trace showed no change. We suspected that pulmonary gas embolism caused sudden decreases in Petco2and Spo2. Immediately, nitrous oxide was discontinued and the patient was placed in a head-down position. Several minutes after the interruption of MCT, Petco2and Spo2value returned to baseline. The patient recovered from anesthesia uneventfully, and there were no significant postoperative complications.
This case prompted us to further investigate whether gas emboli would be produced by localized internal thermal therapy of the liver. With approval from our University Ethical Committee and informed consent, eight patients, classified as ASA physical status 2, undergoing MCT (three male and one female, aged 56–70 yr, 49–70 kg; Microtaze® Model OT-110 M), or RFA (four male, aged 58–77 yr, 49–56 kg; Cool-tip®, RF System, Radionics, Burlington, MA) for hepatocellular carcinoma in a supine or left lateral recumbent position were studied. None of the patients had evidence of esophageal varices. After induction of anesthesia and tracheal intubation with intravenous fentanyl, thiamylal, and vecuronium, a four-chamber view of the heart was monitored using a 5 MHz transducer (PEF-507SB®; Toshiba Medical, Tokyo, Japan) for a transesophageal echocardiogram. Anesthesia was maintained with inhalation of isoflurane or sevoflurane in a 1:1 gas mixture of oxygen and nitrous oxide. The output/frequencies of MCT and RFA were 60 W/2,450 MHz and 70–90 W/350–450 KHz, respectively.
In all eight cases, MCT or RFA resulted in gas bubbles in the right atrium or right ventricle. Bubbles were identified as highly echogenic dots with high mobility. A gas-bubble–like image also appeared in the left atrium and left ventricle in one patient undergoing MCT (fig. 1). The occurrence of gas emboli continued throughout the thermal intervention, then gradually decreased after the end of heating. There were, however, no remarkable changes in the other variables, including Petco2during MCT or RFA.
Venous gas emboli can occur during several types of hepatic interventions, such as electrocauterization, 7argon-enhanced coagulation, 8,9and water jet dissection. 10The present report showed the generation of gas emboli detected by transesophageal echocardiogram during hepatic surgery using MCT or RFA.
MCT has been applied clinically to obtain hemostasis during transection of the hepatic parenchyma and in the treatment of nonresectable liver cancer. Microwave electrodes induce a rapidly alternating electromagnetic field. As water molecules follow the changing polarity of the field, heat is generated within the tissue, resulting in coagulation necrosis and hemostasis. RFA is a thermal technique that causes local tissue destruction by coagulation necrosis. Rather than serving as an external source of tissue heating, radiofrequency waves induce ionic agitation, which results in frictional heat production within the tissue.
The etiology of gas emboli during these interventions is unclear. Although the exact temperature around the electrode is unknown, the electrode tip temperature will reach more than 80°C. 3We speculate that evaporation of gases, including nitrous oxide dissolved in the blood and tissue, is a cause of gas bubbles produced during MCT or RFA. Also, the entrainment of atmospheric air from the opened venous sinuses of the liver may exacerbate gas embolism during laparotomic MCT or RFA.
There have been no clinical reports on harmful embolisms caused by MCT or RFA. However, we make the following recommendations for these procedures: (1) The minimal duration of MCT or RFA application required to achieve tissue necrosis should be selected. (2) Precautions should be taken for harmful gas embolism, particularly in patients with the potential for intracardiac right to left shunt, such as persistent foramen ovale. (3) Appropriate monitors, including transesophageal echocardiography, should be used in cases in which extensive therapy is planned.