Secondary Logo

Journal Logo

Arterial Oxygenation During One-Lung Ventilation

Combined Versus General Anesthesia

Garutti, Ignacio, MD; Quintana, Begona, MD; Olmedilla, Luis, MD; Cruz, Alberto, MD; Barranco, Monica, MD; de Lucas, Elvira Garcia, MD

doi: 10.1213/00000539-199903000-00005
Cardiovascular Anesthesia: Society of Cardiovascular Anesthesiologists

The optimal anesthetic management of patients undergoing thoracotomy for pulmonary resection has not been definitely determined.We evaluated whether general IV anesthesia (propofol-fentanyl) provides superior PaO2 during one-lung ventilation (OLV) compared with thoracic epidural anesthesia (TEA) with supplemental local and general anesthetics. We studied 60 patients who had prolonged periods of OLV for elective thoracic surgery for lung cancer and who were prospectively randomized into two groups. In 30 patients (GA group), fentanyl/propofol/rocuronium anesthesia was used. Another 30 patients (TEA group) were anesthetized with propofol/rocuronium/epidural thoracic bupivacaine 0.5%. A double-lumen endotracheal tube was inserted, and mechanical ventilation with 100% oxygen was used during the entire study. Arterial and venous blood gases were recorded before surgery in a lateral position with two-lung ventilation, 15 and 30 min after OLV (OLV + 15 and OLV + 30, respectively) in all patients. We measured PaO2, venous central oxygen tension, arterial and central venous oxygen saturation, venous admixture percentage (Qs/Qt%), and arterial and central venous oxygen content. The mean values for PaO2 during OLV in the GA group after 15 min (175 mm Hg) and 30 min (182 mm Hg) were significantly (P < 0.05) higher compared with the TEA group (120 and 118 mm Hg, respectively). Furthermore, Qs/Qt% was significantly (P < 0.05) increased in the TEA group during OLV. There were no other significant differences. We conclude that using the TEA regimen is associated with a lower PaO2 and a larger intrapulmonary shunt during OLV than with total IV anesthesia alone. Implications: Sixty patients undergoing elective lung surgery during a prolonged period of intraoperative one-lung ventilation were studied and randomized to receive general IV anesthesia or general IV anesthesia combined with thoracic epidural anesthesia. The arterial oxygenation in the first group was better than that in the second group during one-lung ventilation.

(Anesth Analg 1999;88:494-9)

Service of Anesthesiology and Reanimation, Hospital General Gregorio Maranon, Madrid, Spain.

Accepted for publication December 2, 1998.

Address correspondence and reprint requests to Dr. Ignacio Garutti Martinez, Hospital General Gregorio Maranon, C/Doctor Esquerdo N-46, Madrid 28009, Spain. Address e-mail to

During one-lung ventilation (OLV) with patients in the lateral decubitus position, there is a potential risk of considerable intrapulmonary shunting of deoxygenated pulmonary arterial blood, which may result in hypoxemia. The consequences of an increase in pulmonary vascular resistance (PVR) in the nondependent (nonventilated) lung is to redistribute blood flow to the ventilated dependent lung, thereby preventing PaO (2) from excessively decreasing. This increase in nondependent lung pulmonary vascular resistance is predominantly due to hypoxic pulmonary vasoconstriction (HPV) [1].

Thoracic epidural anesthesia (TEA) with local anesthetics during OLV is increasingly being combined with general anesthesia (GA) in our clinical practice for thoracic surgery. A combination of TEA with GA might maximize the benefits of each form of anesthesia [2]. Furthermore, epidural anesthesia and postoperative epidural analgesia may improve outcome in high-risk patients [3]. Potential disadvantages include the time required to establish epidural anesthesia, intravascular fluid administration needed to avoid hypotension, and the potential for technical complications, such as epidural hematoma.

The effect of intraoperative TEA with local anesthetics on HPV during thoracic surgery and OLV is unclear. The pulmonary vasculature is innervated by the autonomic nervous system, and the sympathetic none is dominant in the pulmonary circulation relative to parasympathetic activity. Theoretically, a TEA-induced sympathectomy might attenuate HPV [4]. However, in one recent experimental study [5], TEA did not affect the primary pulmonary vascular tone, but it improved PaO2 because of enhanced blood flow diversion from the hypoxic lobe.

The purpose of this investigation was to compare oxygenation with a propofol-fentanyl infusion-maintained anesthesia versus propofol infusion combined with thoracic epidural bupivacaine anesthesia during surgery with OLV in humans, using a prospective randomized design.

Back to Top | Article Outline


We studied 60 patients undergoing elective lung surgery during a prolonged period of intraoperative OLV. The study was approved by our ethics committee, and signed, informed consent was obtained from each patient. Exclusion criteria were renal insufficiency (creatinine >1.5 mg/dL), liver dysfunction (aspartate amino transferase >40 U/L, alanine amino transferase >40 U/L), or documented coronary or vascular disease. No patient had a history of myocardial infarction or arrhythmia before the operation. The patients were not premedicated before arrival in the operating room.

For all patients, preoperative assessment included measurements of forced vital capacity (FVC), forced expiratory volume in 1 s (FEV1), these values as a percentage of predicted values (FVC%, FEV1%), and arterial blood gas analysis on the evening before surgery.

The patients were randomized to one of two study groups by lottery: general IV anesthesia (GA group) or general IV anesthesia combined with TEA (TEA group). In the GA group, general anesthesia was induced using fentanyl IV (3 [micro sign]g/kg), midazolam (2-3 mg), and propofol (2 mg/kg); rocuronium (0.6 mg/kg) was given to facilitate intubation of the trachea with a double-lumen endobronchial tube. Anesthesia was maintained with propofol at continuous perfusion (6-7 mg [center dot] kg-1 [center dot] h-1), with increments of fentanyl IV (2 [micro sign]g/kg) to maintain the systolic blood pressure within 15 mm Hg of postinduction values and rocuronium at continuous perfusion (0.5 mg [center dot] kg-1 [center dot] h-1). In the TEA group (combined anesthesia), an epidural catheter was placed at the T6-7 or T7-8 interspace and advanced 3 cm in the epidural space before anesthesic induction. TEA was then induced using an initial 6 to 8-mL dose of plain bupivacaine 0.5%; if necessary, additional increment doses up to 14 mL were administered until a thoracic-sensitive blockade was induced. The level of anesthesia was determined by the loss of pinprick sensation. During the onset of epidural anesthesia, colloids were infused (7 mL/kg); crystalloids (8 mL [center dot] kg-1 [center dot] h-1) were subsequently infused throughout the study (the same rate as the GA group), and when systolic arterial blood pressure decreased to <100 mm Hg, ephedrine was injected in increments of 5 mg. GA was induced using the same method as in GA group. After tracheal intubation, with a double-lumen endobronchial tube, anesthesia was maintained by continuous epidural infusion (6-8 mL/h) of bupivacaine 0.375% and propofol in continuous perfusion (6-7 mg [center dot] kg-1 [center dot] h-1) and rocuronium (0.5 mg [center dot] kg-1 [center dot] h-1) in continuous perfusion.

In both groups, transfusion management and ephedrine administration were based on hematocrit measurements and hemodynamic monitoring and were under the direction of the attending anesthesiologist.

After the induction of anesthesia, an arterial catheter was placed in the radial artery, and a central venous line (two lumens 20 cm long) was introduced via the internal right jugular vein into the right atrium, and its position was confirmed by chest roentgenogram.

After clinical confirmation of correct double-lumen tube placement (by inspection and auscultation) with the patient in both the supine and lateral decubitus positions, ventilation was controlled by using 100% oxygen and a tidal volume of 8-10 mL/kg at a rate to maintain the PaCO2 between 35 and 40 mm Hg. Effective lung isolation was determined by the absence of a leak from the nonventilated lumen of the endobronchial tube. When the pleura was opened, the isolation was confirmed by direct observation of the collapsed nonventilated lung and the absence of a leak from this lung. During OLV, the same tidal volume, respiratory rate, and fraction of inspired oxygen were used and the bronchus to the lung not ventilated upon being excluded and open to atmospheric pressure. Arterial and venous blood gases, heart rate (HR), and mean arterial pressure (MAP) were measured, always with the patient in the lateral position, in three phases: during two-lung ventilation (TLV), 15 min after beginning OLV (OLV + 15), and 30 min after beginning OLV (OLV + 30). These measurements were made before ligation or division of any pulmonary vessels or bronchi.

Blood samples were drawn simultaneously from the distal central venous lumen and arterial catheters and analyzed within 5 min.

The Qs/Qt% at each phase was calculated using the venous admixture equation: Qs/Qt% = (Cc[prime]O2 - CaO2)/Cc[prime]O2 - CvO2) x 100 where the oxygen content of pulmonary capillar blood (Cc[prime]O2) = (Hb x 1.39)SaO (2) + (PaO2 x 0.0031) and Hb = hemoglobin.

PaO2 was calculated using the formula: PaO2 = FIO2 x (Pb-PH (2) O) - (PaCO2/0.8) where Pb = pressure of atmosphere and FIO2 = fraction of inspired O2. CaO2 or CVO2 = oxygen content (mL O2/100 mL blood) (arterial or venous, respectively) was calculated using the formula C(a or v)O2 = (1.34 x Hb x %Sat) + (0.0031 x PO2), but assuming that the SvO2 (mixed venous saturation of oxygen) was equal to the saturation of the blood sample extracted from the right atrial (SvcO2).

Statistical analysis was performed with the aid of a computer program; in all tests, the level of significance was set at 5%. Mean preoperative demographic and laboratory variables of the GA and TEA subjects were compared using Student's t-test for independent samples. The effects of the anesthetic on arterial blood gas and hemodynamic variables were tested by multivariate repeated-measures analysis of variance, using one between-subjects factor (group) and two within-subject factors (anesthetic and number of lungs ventilated). Comparison of PaO2 values at specific times during OLV was by using paired Student's t-test with Bonferroni correction for multiple comparisons.

Back to Top | Article Outline


There were no significant differences between the GA and TEA groups with respect to any of the preoperative data (Table 1). In all the patients, a decrease in PaO2 and an increase in the Qs/Qt% occurred during OLV. These changes were larger in the TEA group compared with the GA group (P < 0.01) (Table 2). PaO2 values were significantly higher after 15 and 30 min OLV in the GA group, more so than the TEA group (P < 0.05). Three patients from the GA group and two from the TEA group presented with values of PaO2 <70 mm Hg during the two measurements conducted in OLV.

Table 1

Table 1

Table 2

Table 2

There were no statistical differences between the two groups in PaCO (2), PvO2, SvCO2, mean airway pressure, HR, CaO2, or CvO2 in the three measurements conducted (Table 2).

In starting OLV, there was a significant decrease in PaO2, CaO2, and CvO2, and an increase in the Qs/Qt% in the two groups of patients. However, there were no significant changes when comparing the values of OLV + 15 with those of OLV + 30.

There were significant differences between the two groups with respect to Qs/Qt% in OLV + 15 and OLV + 30 (Table 2). The two groups presented values of Qs/Qt% similar to each other during the TLV. However, with OLV, there was an increase of these values in both groups: the intrapulmonary shunt increased to 32.1% in the GA group, and the TEA group showed a shunt of 37.4%. When comparing these values between the two groups in OLV + 15 and OLV + 30, there were significant differences (P < 0.05).

Two patients in the GA group and five in the TEA group had a peak Qs/Qt% >or=to40% during OLV. The shunt increased by 76% in the GA group and by 115% in the other group during the OLV (P < 0.05). In analyzing the Qs/Qt% of the patients, depending on the operated side, we observed that during OLV, in the GA group, we obtained values of 29.9% and 34.2% in the operations on the left side and right side, respectively. In the TEA group, these values were 36.7% and 38.3% for the left operated side and right operated side, respectively, during OLV.

There were no complications associated with the TEA technique (epidural hematoma, neurological damage, etc.). In 15 patients, there was a significant decrease in systemic arterial blood pressure with the administration of local anesthetics in the epidural space, which was treated by administering a dose of 10 mg of IV ephedrine. In all these cases, this drug was administered before the start of OLV. In the GA group, ephedrine administration was not required at any time.

Back to Top | Article Outline


OLV creates an obligatory transpulmonary shunt through the atelectatic lung. Passive (gravity and surgical interferences) and active (HPV) mechanisms minimize the diversion of blood flow to the atelectatic lung and prevent the PaO2 from decreasing; however, the most significant diversion of blood flow to the dependent lung is caused by HPV [6].

We demonstrated that when comparing both anesthetic techniques, using combined anesthesia (GA plus TEA with local anesthetics) for thoracic surgery produces a larger decrease in the PaO2 and a larger increase in the intrapulmonary shunt during OLV than does IV GA.

In a preliminary study, the effects of TEA and GA on PaO2 and intrapulmonary shunt were compared. The values for PaO2 were decreased by adding local anesthetics by means of a thoracic epidural catheter; however, with so few patients in the group (n = 8), it is impossible to extract any definite conclusions for statistic purposes [7].

The reason for the decrease PaO2 is uncertain: the pulmonary vasculature is innervated by the autonomic nervous system. Stimulation of sympathetic nerves to the lung causes an increase in PVR produced by activation of alpha-receptors in the pulmonary vascular bed. The mediator released from sympathetic nerve endings is norepinephrine [8]. Blockade of the sympathetic nervous system with alpha-adrenergic antagonists or beta-adrenergic agonists attenuates HPV, whereas the beta-adrenergic antagonists increase this response.

Perhaps block of the activity of the thoracic sympathetic over the vascular pulmonary response is a factor. In these studies, TEA did not affect the primary pulmonary vascular tone during OLV, but it slightly enhanced the diversion of blood flow from the hypoxic lobe to the other well oxygenated areas of the lung [5]. However, Kazemi et al. [9] reported a reduction in the hypoxic response of the lung after sympathectomy. The problem is that most these studies were not performed under the same conditions as those in the present study (e.g., anesthetized patients, lateral decubitus position, and atelectatic lung).

Other factors that could decrease PaO2 are the cardiovascular and hemodynamic effects of TEA: decrease in HR, MAP, stroke volume, and cardiac output (CO) due to blockade of the sympathetic nervous system. Furthermore, the systemic effects of the absorption of the local anesthetics can contribute to circulatory changes, such as a decrease in CO [4,10]. We do not believe that the best oxygenation during OLV in the GA group could be due to a decrease in CO, because this mechanism fundamentally affects the HPV through variations in the SvO2 and PvO2. In our patients, we observed values very similar in the SvCO2 and PvO2 in the two groups of patients during the study. However, there were significant differences in the values of PaO2 and the Qs/Qt% in OLV + 15 and OLV + 30.

When the nondependent lung is ventilated in lateral decubitus position, it receives approximately 40% pulmonary blood flow [11]. Our patients (both groups together) had an average shunt of 18.8% during TLV with the patient in the lateral decubitus position. If this shunt was equally distributed between the nondependent and dependent lungs and if blood flow to the atelectatic nondependent lung was not acutely decreased, then the shunt during OLV would have been 49.4% (18.8/2% + 40%) for all our patients (because the same number of right- and left-side lung operations were performed). However, in the GA group, the shunt was 31%, versus 39.5% in the TEA group. The most likely mechanism of the acute decrease in blood flow to the atelectasic nondependent lung is HPV [12]. If so, HPV was better conserved in the GA group. In the TEA group, the lower PaO2 and higher shunt values during OLV are consistent with a greater depression of HPV, probably due to sympathectomy and/or systemic absorption of local anesthetics.

Another unlikely possible mechanism for acute decrease in nondependent lung blood flow during clinical OLV could be surgical interference. However, we had the cooperation of our surgeons to ensure that no manipulation, retraction, or clamping of the other lung occurred during the measurements. We felt that the PaCO2 was well controlled. The patients from the TEA group showed values slightly greater than those of the GA group, although not statistically significant. Yet, we know that hypercapnia during OLV seems to act as a vasoconstrictor by selectively increasing ventilated lung pulmonary vascular resistance (enhanced directly regional HPV) [13]. High airway pressures in the dependent lung may counteract HPV in the non-dependent lung by diverting blood flow away from the ventilated lung, thereby increasing the pulmonary shunt fraction [14]. However, the average values of airway pressures in both groups are similar.

HPV is a primary regulator of blood flow distribution in the atelectatic lung, of which the alveolar oxygen tension is the primary stimulus, but with atelectasis, the oxygen tension of most lung tissue approaches the mixed venous oxygen tension. Domino et al. [15] showed that when PvO2 was normal (46 +/- 2 mm Hg) or lower, HPV occurred, and approximately 50% of the blood flow was diverted away from the atelectatic lung; however, when PvO2 was high (i.e., 100-140 mm Hg), HPV was inhibited. For ethical reasons, venous samples were drawn from the right atrium in this study to obviate the need for pulmonary artery catheterization. Antman et al. [16] indicated that the average difference between pulmonary artery and right atrial oxygen concentration was 0.34% +/- 2.5 SD. Berridge et al. [17] suggest that ScvO2 is a good estimate of SvO2, even when the exact position of the central venous catheter is unknown. In our patients, the values of PvO2, as well as those of SvCO2, stayed in a similar range in the two groups of patients, thereby leading us to think that this could not be the cause of the poor oxygenation and higher shunt of the TEA group. Finally, the administration of ephedrine was only necessary in the TEA group. This explains the similarity of the compared values of HR and MAP in the two groups, but it does not explain poorer oxygenation, because ephedrine seems to produce an increase in PaO2 without altering the intrapulmonary shunt in OLV during thoracic surgery [18].

We conclude that vasodilation by sympathetic blockade counteracts HPV and thereby produces a larger shunt and a decrease in oxygenation during the OLV. Although TEA likely facilitates early extubation of patients undergoing thoracic surgery, we cannot confirm this conclusion. The anesthesiologist who extubated the trachea knew which anesthetic technique was used, which could cause bias.

In summary, TEA during OLV with the patient in the lateral decubitus position increases the intrapulmonary shunt and decreases the PaO2. We conclude that TEA cannot be recommended in patients undergoing OLV.

Back to Top | Article Outline


1. Cutaia M, Rounds S. Hypoxic pulmonary vasoconstriction: physiologic significance, mechanism, and clinical relevance. Chest 1990;97:706-18.
2. Temeck BK, Schafer PW, Park WY, Harmon JW. Epidural anesthesia in patients undergoing thoracic surgery. Arch Surg 1989;124:415-8.
3. Yeager MP, Glass DD, Neff RK, Brinck-Johnsen T. Epidural anesthesia and analgesia in high-risk surgical patients. Anesthesiology 1987;66:729-36.
4. O'Connor CJ. Thoracic epidural analgesia: physiologic effects and clinical applications. J Cardiothorac Anesth 1993;7:595-609.
5. Ishibe Y, Shiokawa Y, Umeda T, et al. The effect of thoracic epidural anesthesia on hypoxic pulmonary vasoconstriction in dogs: an analysis of the pressure-flow curve. Anesth Analg 1996;82:1049-55.
6. Benumof JL. One-lung ventilation and hypoxic pulmonary vasoconstriction: implications for anesthesic management. Anesth Analg 1985;64:821-3.
7. Maseda F, Vilchez E, Del Campo JM, et al. Hypoxic pulmonary vasoconstriction during single lung ventilation in the lateral decubitus position and thoracic anaesthesia [abstract]. Br J Anaesth 1995;74:A152.
8. Kadowitz PJ, Hyman AL. Effect of sympathetic nerve stimulation on pulmonary vascular resistance in the dog. Circ Res 1973;32:221-7.
9. Kazemi H, Bruecke PE, Parsons ES. Role of the autonomic nervous system in the hypoxic response of the pulmonary vascular bed. Respir Physiol 1972;15:245-8.
10. Wattwil M, Sundberg A, Arvill A, Lennquist C. Circulatory changes during high thoracic epidural anaesthesia-influence of sympathetic block and systemic effect of the local anaesthetic. Acta Anaesthesiol Scand 1985;29:849-55.
11. Benumof JL. Anesthesia for thoracic surgery. Philadelphia, WB Saunders, 1987.
12. Benumof JL. Mechanism of decreased blood flow to alectatic lung. J Appl Physiol 1979;46:1047-8.
13. Westbrook JL, Sykes MK. Peroperative arterial hypoxaemia: the interaction between intrapulmonary shunt and cardiac output. Anaesthesia 1992;47:307-10.
14. Hambraeus-Jonzon K, Bindslev L, Mellgard AJ, Hedenstierna G. Hypoxic pulmonary vasoconstriction in human lungs: a stimulus-response study. Anesthesiology 1997;86:308-15.
15. Domino KB, Wetstein L, Glasser SA, et al. Influence of mixed venous oxygen tension (PvO2) on blood flow to atetectatic lung. Anesthesiology 1983;59:428-34.
16. Antman EM, March JD, Green LH, Grossman W. Blood oxygen measurements in the assessment of intracardiac left to right shunts: a critical appraisal of ethodology. Am J Cardiol 1980;46:265-71.
17. Berridge JC. Influence of cardiac output on the correlation between mixed venous and central venous oxygen saturation. Br J Anaesth 1992;69:409-10.
18. Tanaka M, Dohi S. The effects of ephedrine and phenilephrine on arterial partial pressure of oxygen during one-lung ventilation. Masui 1994;43:1124-9.
© 1999 International Anesthesia Research Society