ARTERIAL OXYGENATION DURING THORACIC SURGERY

Abstract

Introduction. Arterial oxygenation (PaO2) during anesthesia for thoracic surgery remains a clinically important problem. Clinical studies suggest that PaO2 during one-lung ventilation (OLV) with isoflurane is superior to halothane [1] or enflurane [2] and equivalent to total intravenous anesthesia [3], but these have not been compared to sevoflurane. This study compared gas exchange and hemodynamics before and during stable OLV with isoflurane and sevoflurane in a randomized crossover design. Methods. Twenty consenting patients having prolonged OLV for thoracic surgery were studied with IRB approval. Patients were randomized into 2 groups and received a standardized anesthetic induction with propofol, vecuronium and sufentanil. In group 1 isoflurane in oxygen was titrated to maintain a FET isoflurane - 1 MAC (1.2%). Arterial and mixed-venous blood gases were drawn and thermodilation cardiac output measured during two lung ventilation and every 10 min of OLV with the chest open for at least 30 min, or until the PaO2 change was < 25 mmHg. Isoflurane was then discontinued and sevoflurane titrated to maintain FET 1 MAC (1.7%). Blood gases and hemodynamics were measured q. 10 min for a further 30 min, monitoring the FET of both agents simultaneously. The study was completed before surgical manipulation of the lung or pulmonary vessels. Group 2 subjects followed an identical protocol except that the order of isoflurane and sevoflurane administration was reversed. The effects of anesthetic sequence and agents on gas exchange, hemodynamics and recovery were tested by repeated measures ANOVA. Results. There were no significant differences between the groups for baseline or demographic data. During two-lung ventilation there were significant differences between groups, mean (+ SD) PaO2 was higher: 474 (+/- 58) mmHg and shunt (Qs/Qt) was lower 0.14 (+/-.05) with sevoflurane than isoflurane: 418 (+/- 49) mmHg (p =.03) and 0.23 (+/-.08) (P <.01). There were no significant differences due to anesthetic sequence or agent during OLV, mean PaO2 and shunt after 30 min of isoflurane were: 200 (+/- 111) mmHg and 0.34 (+/-.10) and after 30 min sevoflurane: 178 (+/- 77) mmHg and 0.33 (+/-.08). Spontaneous eye opening after termination of anesthesia was faster following sevoflurane: 7.1 (+/- 5.6) min than isoflurane: 13.7 (+/- 6.9) (p =.03). Discussion. During OLV no significant differences were discovered between 1 MAC isoflurane and 1 MAC sevoflurane with respect to PaO2, shunt, cardiac output or other hemodynamics. The decreased shunt and higher PaO2 during two-lung ventilation and more rapid recovery with sevoflurane may offer some clinical advantages.