Although gas embolism is a rare complication, cerebral infarction by gas embolism is particularly rare. According to a systematic literature review, 3 cases (0.2%) of gas embolism were reported among 1262 patients who underwent laparoscopic major hepatectomy . The reported mortality rate of CO2 embolism is 28% . Once CO2 enters the systemic venous system, it is transported to the lungs and CO2 embolism occurs, resulting in cardiac arrhythmias, pulmonary hypertension, and hypotension . In comparison with general laparoscopic surgery, the hepatic vein could be sufficiently exposed on the surface to be transected during anatomical resection in major hepatectomy [8, 10]. Thus, CO2 embolism might more frequently occur in laparoscopic hepatectomy than in the other laparoscopic surgeries.
Paradoxical embolism provokes systemic embolism caused by the presence of an intracardiac defect, intrapulmonary arteriovenous shunt, or pulmonary fistula [2, 6, 7, 11]. Intrapulmonary shunts may occur in patients with direct arteriovenous communications and dilated pre-capillaries in the lungs secondary to liver cirrhosis or pulmonary arteriovenous malformation . Because almost all cerebral gas embolisms are diagnosed after observation of delayed recovery from general anesthesia or impaired consciousness, the diagnosis of paradoxical embolism during surgery is difficult [1, 7].
A patent foramen ovale (PFO) is found in approximately 25% of healthy individuals . However, a PFO is usually closed under normal physiological conditions. Thus, detection of a small PFO before and after surgery is difficult because opening of the PFO and right-to-left shunting mainly occur when the right atrial pressure exceeds the left atrial pressure . Transesophageal echocardiography (TEE) is considered the gold standard for diagnosis because it allows for direct anatomic visualization of shunting . Actually, TEE can reveal the existence of the PFO in approximately 24% of patients in contrast to transthoracic echocardiography (15% of patients). However, an opened PFO is not necessarily observed by TEE with manual positive-pressure ventilation during surgery. In the present case, we should have examined the patient for an air embolism using TEE when the ST change developed during surgery.
The reason for the brain infarction in the present case is unclear, but we considered three possibilities. The first possibility is that the right atrial pressure exceeded the left atrial pressure secondary to the CO2 embolism caused right-to-left shunting through an open PFO. This possibility could also explain the ST change in the electrocardiogram. The second possibility is the development of low blood pressure. The minimum intraoperative blood pressure was 51/24 mmHg and the duration of time for which the systolic pressure was < 60 mmHg was about 3 min. However, it seems that this hypotension period was not clinically important. The third possibility is the development of an intrapulmonary shunt caused by liver cirrhosis. Such shunts develop in 15 to 45% of patients with liver cirrhosis . Kawahara et al.  reported the occurrence of cerebral infarction during laparoscopic surgery. They concluded that the cause was an intrapulmonary shunt because an intracardiac shunt was not observed by TEE.
The mechanism of ST change in this case could have been a gas embolism or coronary spasm. However, the present report indicates that an ST change on the electrocardiogram could be helpful when paradoxical embolism is suspected. Additionally, if we had acted quickly to reduce the risk of paradoxical air embolism (i.e., moving the patient to the Trendelenburg position, initiating TEE, changing the procedure to open surgery, or suctioning air from the right atrium via central venous catheter if present) when the ST change developed in this case, the cerebral infarction might have been prevented. Thus, we suggest preoperative TTE examination for detection of PFO and intraoperative monitoring of the bispectral index or regional oxygen saturation which could possibly detect gas embolism . TEE is reportedly the most sensitive technique for detecting intracardiac bubbles [2, 5]. However, we believe that TEE should be used for high-risk patients who are expected to have a large amount of bleeding. Furthermore, an adequate central venous pressure should be maintained, and any increase in the intra-abdominal pressure during laparoscopic hepatectomy should be addressed.