The goal of this randomized clinical trial is to compare a liberal versus restrictive oxygen supply (fraction of inspired oxygen, FiO2) strategy in patients scheduled for thoracic surgery requiring one-lung ventilation during lung isolation. The primary and secondary outcome parameters are:

• oxygenation of the blood after 30 minutes of one-lung ventilation, assessed by PaO2/FiO2 ratio
• time to lung collapse after start of one-lung ventilation

Participants in the control goup will receive an oxygen content of 100% before lung isolation, which will be subsequently decreased to achieve normoxia or mild hyperoxia (PaO2 of 75-120 mmHg).

The intervention group will receive the previous, during two-lung ventilation set, oxygen content and after lung isolation oxygen supply will be increased to secure adequate oxygenation of the blood (PaO2 75-120 mmHg) during one-lung ventilation.

The investigators hypothesize, that a higher fraction of inspired oxygen may impede hypoxic pulmonary vasoconstriction of the collapsed lung and thus decrease overall oxygenation performance during one-lung ventilation. Secondary endpoint will be the time to lung collapse, as a lower fraction of inspired oxygen and thus a higher nitrogen content may impede lung collapse.

After anesthesia induction and securing the airway with a double lumen tracheal tube, the patient will be ventilated with pressure-controlled ventilation (PCV) and the following settings during two-lung ventilation (TLV): positive end-expiratory pressure of 5 cmH2O, peak pressure set to achieve a tidal volume of 6-8 ml/kg, respiratory rate set to achieve normocapnia, I:E ratio set to 1:1.5, FiO2 adjusted to achieve normoxia. After changing to lateral position, the patient will be randomized to one of the following group:

• Decremental FiO2 titration: FiO2 will be set to 1.0 five minutes before lung isolation and reduced consecutively during one-lung ventilation (OLV) according to paO2 measurement obtained from arterial blood gas samples to achieve normoxia (paO2 of 75-120 mmHg).
• Incremental FiO2 titration: FiO2 will be maintained to secure normoxia during the entire surgery. This means, that the previous FiO2 setting during two-lung ventilation will be continued and after OLV initiation the FiO2 has to be adjusted, to secure normoxia assessed by continuous SpO2 measurement (SpO2 of 92-96 %) and paO2 measurement obtained from arterial blood gas samples (paO2 of 75-120 mmHg).

During OLV the applied tidal volume will be reduced to 4-6 ml/kg by an appropriate adjustment of peak pressure, otherwise the ventilator settings will not be changed. After 30 minutes of OLV the intervention period ends and because in both groups oxygenation will be adjusted to achieve a PaO2 of 75-120 mmHg during OLV, the routine anesthetic regime of both groups will not differ in the further course of thoracic surgery.

Inclusion Criteria:

• Male and female subjects ≥ 18 years
• Elective thoracic surgery requiring OLV
• American Society of Anesthesiologists physical status classification I-III
• Written informed consent

Exclusion Criteria:

• Emergency surgery
• Female subjects known to be pregnant
• Known participation in another interventional clinical trial
• Empyema evacuation or signs of pulmonary infection

Study participants will be randomized after induction of general anesthesia and thus are not aware of the treatment arm, which solely differs at the time of lung isolation.