<h3>Introduction</h3> Patients with coronary artery disease (CAD) undergoing general anesthesia are at risk of perioperative complications. Hyperoxia is a known coronary vasoconstrictor. However, it is also a pulmonary vasodilator, reducing right ventricular afterload. Thus, there may be competing effects of hyperoxia on right ventricular function. In this study we investigated the effects of hyperoxia and normoxia on right ventricular (RV) function assessed by 3D transesophageal echocardiography (TEE) during general anesthesia in CAD patients before elective coronary artery bypass graft surgery. <h3>Methods</h3> In this randomized clinical trial study participants (n=106) were prospectively recruited using a crossover design. In anesthetized patients the fraction of inspired oxygen (FiO2) was titrated to a normoxic state (FiO2=0.3) and a hyperoxic state (FiO2=0.8). At both states TEE images were acquired to assess RV ejection fraction (RVEF) and global longitudinal free-wall strain (RVGLS). <h3>Results</h3> There was no difference between normoxia and hyperoxia in RVEF (46±6% vs. 45±8, p=0.504) nor in RVGLS (-22.4±4.8 vs. -21.9±4.3, p=0.352). However, RV function improved and worsened in some patients. ROC analysis (Figure A) shows that RVGLS at normoxia can better predict, which patients will worsen with excess oxygen. With a cut-off of -20% for RVGLS (Sensitivity: 87%, Specificity: 49%), it was demonstrated that hyperoxia was beneficial for patients with a poor strain at normoxia (>-20%) but detrimental for those with normal strain at normoxia (<-20%, Figure B), while RVEF with a cut-off of 40% (Sens: 88%, Spec: 33%) was unable to predict this response (Figure C.). <h3>Conclusion</h3> In CAD patients undergoing general anaesthesia, hyperoxia has heterogenous effects on the RV and the potential detrimental effects of hyperoxia were best predicted by RVGLS. Intraoperative strain analysis might be a tool to target oxygen levels based on individual needs.
Read full abstract