Hypoxia during one lung ventilation in thoracic surgery

Background. The technique of one lung ventilation (OLV) is used with the purpose of achieving isolation of the diseased lung being operated upon, using a doublelumen endobronchial tube. Thoracic surgical procedures which are performed in the lateral decubitus position, nowadays could not be imagined without OLV. In spite of advantages regarding surgical exposure, OLV is associated with serious respiratory impairment. Hypoxemia is considered to be the most important challenge during OLV. The goal of this study was to establish the magnitude of intrapulmonary shunt, as well as the immensity of hypoxia during general anesthesia with OLV. Materials and Methods. In this prospective interventional clinical study thirty patients were enrolled who underwent elective thoracic surgery with a prolonged period of OLV. The patients received balanced general anesthesia with fentanyl/propofol/ rocuronium. A double-lumen endobronchial tube was inserted in all patients, and mechanical ventilation with 50% oxygen in air was used during the entire study. Arterial blood gases were recorded in a lateral decubitus position with two-lung ventilation, at the beginning of OLV (OLV 0) and at 10 and 30 min. (OLV 10, OLV 30, respectively) after initiating OLV in all patients. Standard monitoring procedures were used. Arterial oxygenation (PaO2), arterial oxygen saturation (SaO2) and venous admixture percentage intrapulmonary shunt (Qs/Qt %) were measured, as well as mean arterial pressure and heart rate during the same time intervals. For the purpose of this study, the quantitative value of Qs/Qt% was mathematically calculated using the blood gas analyser AVL Compact 3. A p value <0.05 was taken to be statistically significant. Results. When OLV was instituted, arterial oxygenation decreased, whereas Qs/Qt% increased, about 10 min. after commencement, with improvement of oxygenation approximately half an hour afterwards. A statistically relevant difference (p<0.05) occurred in PaO2, SaO2 and Qs/Qt at the different time points. Conclusion. Hypoxia during OLV, with an increase in Qs/Qt, usually occurs after 10 min. of its initiation. After 30 min, the values of the Qs/Qt ratio regularly return to normal levels.


INTRODUCTION
In order to achieve collapse of the affected lung during thoracic surgical procedures, a technique of ventilation of one lung (OLV) was introduced by inserting a double lumen endobronchial tube.This allows isolation of the dependent lung from the affected, independent lung, which is on top, thus preventing contamination of the healthy lung.On the other hand, the collapse of the affected lung causes serious functional respiratory disorders that require special compensatory measures to avoid hypoxemia.It should be emphasized that maintaining optimum oxygenation is crucial for preventing cellular hypoxia.(1)(2)(3) During OLV with the patient in the lateral decubitus position, there is a potential risk of significant intrapulmonary shunting of deoxygenated pulmonary arterial blood, which may result in hypoxemia.A consequence of the increase in pulmonary vascular resistance (PVR) in the independent (unventilated) lung, predominantly as a result of activated hypoxic pulmonary vasoconstriction (HPV), is a redistribution of blood flow in the ventilated dependent lung.Thus, an excessive drop of partial pressure of oxygen in arterial blood (PaO2) is prevented.(4)(5)(6)(7)(8) Intrapulmonary shunt is the main cause of hypoxemia in OLV, although alveoli with low ventilation/perfusion coupling (Va/ Qt) in the dependent lung contribute to that.In addition, the blood that goes into the upper lung cannot take oxygen, so it keeps the mixed venous composition, which is poorly oxygenated.It is mixed with oxygenated blood in the left atrium of the heart, creating the so-called venous admixture and reducing PaO2.Venous admixture and intrapulmonary shunt (Qs/ Qt) are often used as synonyms.Venous mixture increases from values of approximately 10-15% in ventilation of the two lungs (TLV) to 30-40% in OLV.In most patients, PaO2 can have values in the range of 9-16 kPa using a fraction of inspired oxygen (FiO2) 50-100%.Hypoxic pulmonary vasoconstriction (HPV) is a compensatory mechanism by which the pulmonary blood flow is diverted away from hypoxic/collapsed areas of the lung.This should improve oxygenation during OLV.Volatile anesthetics directly depress HPV, but they also amplify it by decreasing cardiac output (CO).Therefore, there is usually no change in the response to HPV when using volatile anesthetics during thoracotomy and OLV.Intravenous anesthetics, such as propofol, do not inhibit HPV and should improve arterial oxygenation in OLV.There is some evidence in support of this.(9)(10)(11)(12)(13)(14)(15)(16) The purpose of the study was to determine the magnitude of the intrapulmonary shunt and hypoxia during general anesthesia and OLV in thoracic surgery.

MATERIAL AND METHODS
This study was part of a larger prospective clinical trial designed to investigate the influence of thoracic epidural anesthesia on intrapulmonary shunt during OLV in thoracic surgery.It was conducted at the University clinic of anesthesia, reanimation and intensive therapy and the University clinic of thoracic-vascular surgery, University Clinical Centre in Skopje, with the collaboration from our colleagues from the Department of anesthesiology and perioperative intensive therapy from the University Medical Centre in Ljubljana.The study was performed in accordance with the Declaration of Helsinki and after receiving approval from the Ethics Committee of the Medical Faculty in Skopje.Each patient gave written informed consent before being included in the study.The study included 30 patients undergoing elective lung surgery under general anesthesia with employment of OLV (thoracotomy with lung resection -pulmectomy, bi-lobectomy, lobectomy, segmentectomy) or video assisted thoracoscopy (VATS).Relative Shunt volume -The value Qs/Qt is the fraction of venous blood that remains un-oxygenated after traveling from the right side of the heart to the left side of the heart.This fraction includes the effects of true shunts (i.e., anatomic shunts and true capillary shunts) along with the effects of a ventilation-perfusion mismatch.

Calculation of ctO2 ṽ:
With co-oximetry data: ctO2a = (ctHba / ctHb ṽ) ctO2ṽ + 0.00314 PṽO2 ml/dl Without co-oximetry data and with measured mixed venous PO2-values, the equations (13,15) are used for the calculation of ctO2 ṽ -value.Without co-oximetry data and without measured mixed venous PO2-values, the equation ( 15) is used for the calculation of ctO2 ṽ -value The above described relations are only available for a body-temperature of 37 °C.
If the patient's temperature is other than 37 °C, the numerical results are senseless and therefore are not printed.( 17)

Estimates of Qs/Qt
Approximation by the clinical shunt equation to Qs/Qt is provided by estimating the normal arterial to venous oxygen content difference.
Whichever method is used to estimate the intrapulmonary shunt, it should be noted that the value of estimating the shunt is in its ability to help identify changes in lung dysfunction.A change in intrapulmonary shunt can indicate an improvement in the patient's condition if the shunt is decreasing.If the shunt is, however, increasing, the clinical condition may be deteriorating.The use of intrapulmonary shunt estimates, allows the clinician to more routinely analyze changes in the patient's pulmonary status.Without estimating the intrapulmonary shunt, the clinician is limited in understanding how much of a discrepancy exists between how well the lungs are presently oxygenating the blood versus how well they should be oxygenating the blood.This understanding is important based on the levels of clinical support being given, that is increased FiO2, PEEP and mechanical ventilation.In addition, use of shunt estimates helps the clinician avoid simply looking at the PaO2 level.The shunt estimates encourage at least simultaneous analysis of FiO2 and PaO2.It is important to remember that measurement of intrapulmonary shunt does not identify which condition exists.It only reveals the extent of the pulmonary dysfunction induced by the clinical condition.As such, the intrapulmonary shunt is not a diagnostic tool but is used in the assessment of changes in lung function.(18) The 'rationale' behind Qs/Qt calculation
Among patients in the study, 76.66% males and 23.3% females were registered.The difference between sexes is statistically significant for p = 0.003 (

DISCUSSION
OLV creates an obligatory transpulmonary shunt through the collapsed lung.Passive (gravity and surgical manipulation) and active (HPV) mechanisms minimize the diversion of blood flow to the collapsed lung and prevent the excessive fall of PaO2; however, the most significant diversion of blood flow to the dependent lung is caused by HPV. ( 20) Hurford et al. in a study (21) tested the hypothesis that intraoperative hypoxemia during LV is likely to happened when there is a greater preoperative pulmonary blood flow in the operated lung.Their study examined 30 patients who underwent thoracic surgical procedures in the lateral decubitus position using OLV in whom a preoperative ventilation-perfusion scan was performed.The percentage of blood flow in the operated lung seen on preoperative perfusion scan correlated inversely with PaO2 after 10 min. of the start of OLV (r =-.72).When the percentage of blood flow in the operated lung on the preoperative scan was greater than 45%, likelihood of hypoxemia (PaO2<75 mm Hg) was greater.Because in these patients preoperative regional ventilation was equivalent to regional perfusion, the percentage of preoperative ventilation also correlated inversely with PaO2 after 10 min. of the commencement of OLV (r =-.73).Preoperative arterial blood gas analysis, pulmonary function tests, nor lung volumes did not correlate with oxygenation during OLV.This contradicts the results of Slinger and colleagues.(22) In their study they found that an equation with three variables [PaO2 during intraoperative two lung ventilation (TLV) in lateral decubitus position, side of operation and preoperative value of rela-tionship forced expiratory volume in 1st second/vital capacity (FEV1/VC), can be used to predict (p =.73) PaO2 during OLV using application of continuous positive airway pressure (CPAP) to the unventilated lung.However, Katz and colleagues (23) agree with the findings of Hurford and associates ( 21) that routine preoperative gas analysis of arterial blood and pulmonary function tests do not accurately predict which patients are at risk of hypoxemia during OLV.
The results from our study confirm that preoperative blood gas analysis obtained from arterial blood, as well as FVC (forced vital capacity) and FEV1 cannot be taken as conclusive evidence that a particular patient will develop a greater or lesser degree of hypoxia during OLV.Statistically significant differences obtained for PaO2, SaO2 and Qs/Qt, showed that after a certain time from the commencement of OLV (10 min.),hypoxia develops with a decline in the values of PaO2 and SaO2, as well as an increase in the value of the intrapulmonary shunt, which is on the other hand followed also by a rise of mean arterial pressure.Repeated reduction in Qs/Qt in the fourth measurement (T3), suggests the development of HPV in this period and reduction of the shunt fraction 30 min.from the initiation of OLV in the lateral decubitus position during thoracic surgery.Nonetheless, our study has some limitations.Namely, CO and PvO2 (partial pressure of oxygen in mixed venous blood) which are important factors for assessment of the impact of HPV on oxygenation, were not measured.Instead, we have tried to make some analogy between hemodynamic changes in our patients with the alterations of oxygenation status by using the values of MAP obtained by invasive arterial pressure measurement.

CONCLUSION
In patients undergoing OLV during general anesthesia, the development of hypoxia occurs with a decline in PaO2 and increase in the value of intrapulmonary shunt after a certain period of time following commencement of OLV (10 min.).This is followed by a return of PaO2 and Qs/Qt to values close to normal (30 min. of OLV) due to the development of compensatory mechanisms (HPV).

Table 1 .
Distribution of patients by sex

Table 2 .
Distribution of patients by age and weight

Table 3 .
Distribution by ASA status

Table 4 .
Average values of MAP and HR during the four measurements

Table 5 .
t-Test for MAP: Two-Sample Assuming Equal Variances

Table 6 .
Post hoc test for MAP, p value from two sample t-test

Table 7 .
Average values of the parameters of intraoperative arterial blood gas status