Gas embolism during laparoscopic liver resection in a pig model: frequency and severity.

BACKGROUND: Laparoscopic liver surgery is evolving rapidly. Carbon dioxide embolism is a potential complication. The aim of this work was to study the frequency and severity of gas embolism (GE) during laparoscopic liver lobe resection in a pig model and the resulting cardiovascular and respiratory changes.
METHODS: Fifteen anaesthetized piglets underwent laparoscopic left liver lobe resection. Haemodynamic and respiratory variables were monitored, including systemic and pulmonary arterial pressures, end-tidal CO2, and pulmonary dead space. Online blood gas monitoring and a transoesophageal echocardiography (TOE) were used. GE was graded semi-quantitatively as grade 0 (none), grade 1 (minor), or grade 2 (major), depending on the TOE results.
RESULTS: In 10 of 15 piglets, GE occurred. In total, 33 separate episodes of GE were recorded. All 13 episodes of grade 2 and three of grade 1 were serious enough to cause mainly respiratory, but also haemodynamic effects. Mostly, grade 1 GE caused only minor respiratory or haemodynamic changes. Most variables were affected during grade 2 GE; the most important were Pa(o(2)), Pa(co(2)), end-tidal CO2, Vd/Vt, and mean pulmonary arterial pressure.
CONCLUSIONS: GE occurred frequently during laparoscopic liver resection in this experimental study. Approximately half of the embolisms were serious enough to cause respiratory or haemodynamic disturbances or both. Pending further human studies, a combination of several monitoring techniques, with narrow limits for the alarm settings, will ensure correct interpretation of the complex physiological response to GE and reveal it early enough to alert the anaesthetist and the surgeon to the ongoing problem.