Arterial hypoxemia is a serious complication of one-lung ventilation (OLV), with an incidence of <1% in patients during thoracic surgery (1). It is suggested that pressure-controlled ventilation (PCV) results in improved arterial oxygenation during OLV compared with the more widely used volume-controlled ventilation (VCV) (2). This has been attributed to the decelerating inspiratory flow delivery method used during PCV, whereby high initial flow rates are delivered to quickly achieve and maintain the set inspiratory pressure followed by rapidly decelerating flow (3). These high initial rates of flow lead to a more rapid alveolar inflation (4). These mechanical effects of PCV allegedly allow a more homogeneous distribution of ventilation to the lung, leading to better ventilation/perfusion matching (V/Q) (4). The use of PCV during OLV was found to reduce intrapulmonary shunt and peak airway pressures, limiting the risk for barotrauma (2).
Despite these potential benefits, the use of PCV in patients with adult respiratory distress syndrome (ARDS) did not result in improved arterial oxygenation compared with VCV (4–7). Moreover, a study on induced acute lung injury (ALI) in pigs showed an improvement in oxygenation during VCV when compared with PCV. This improvement in Pao2 with VCV was due to a more favorable ventilation/perfusion distribution (8). In addition, the high peak inspiratory flow rates with PCV may possibly aggravate lung injury with sheer forces during inspiration on the airways and alveoli than lower peak inspiratory flow of VCV (9). Nonetheless, the potential benefits versus limitations of ventilation mode during OLV have not been clearly defined. The aim of this study was to investigate whether PCV results in improved arterial oxygenation compared with VCV during OLV in patients undergoing thoracic surgery.
Approval of the study protocol was obtained from the local ethics committee, and all patients gave written informed consent before inclusion. Inclusion criteria for the study were patients ASA physical status II–III undergoing elective open thoracotomy in the lateral position with at least 1 h of OLV. Exclusion criteria were uncompensated cardiac disease or hemodynamically significant arrhythmias.
All patients were premedicated with midazolam 5 mg IV on arrival to the operating room. A paravertebral block was performed at the T6 level with 15 mL of 0.5% bupivacaine via an inserted catheter. General anesthesia was induced with fentanyl 5 μg/kg, propofol 1.5–2 mg/kg, and atracurium 0.5 mg/kg IV was given for muscle relaxation. Anesthesia was maintained with a continuous infusion of propofol 6–10 mg/kg/h, with supplemental fentanyl and atracurium. No volatile anesthetics were used. Heart rate, direct arterial blood pressure, the electrocardiogram, arterial oxygen saturation (Sao2), and end-tidal carbon dioxide tension (ETco2) were continuously monitored (Datex Engstrom, Finland).
The trachea was intubated with a left-sided double-lumen tube (DLT) (Broncho-part, Rush, Kermen, Germany: 41 F for male and 37 F for female patients). Positioning of the DLT was confirmed with fiberoptic bronchoscopy before and after the patient was placed in the lateral position. Lung separation was confirmed by placing the sidestream of the capnograph sensor into each lumen of the DLT while maintaining ventilation through the other lumen. During OLV, the lumen of the nonventilated lung was left open to air.
The patients' lungs were ventilated with a VCV (Vector ventilator, Temel, Valencia, Spain) using the following variables: inspired oxygen fraction of 1.0, tidal volume (VT) 9 mL/kg, and respiratory rate 12 breaths/min, then adjusted to maintain the ETco2 tension between 30 and 35 mm Hg. During VCV the inspiratory time was 33% with end-inspiratory pause of 0.9 s. No external positive end expiratory pressure was applied throughout the entire study. Upon initiation of OLV the patients were randomly allocated to one of two groups (random numbers generated by the Excel computer program, Microsoft, Redmont, WA). Group A: OLV initiated with VCV (OLV-VCV) without modifying respiratory variables (VT 9 mL/kg). After 30 min the ventilator was switched to PCV (decelerating flow) and the inspiratory pressure was adjusted to obtain the same Vt as during VCV. Group B: OLV initiated with PCV with an inspiratory pressure that provided the same Vt as during two-lung ventilation. After 30 min the ventilator was switched to VCV with a Vt 9 mL/kg. Arterial Pao2, Paco2, pH, arterial blood pressure, peak inspiratory pressure (Ppeak), mean inspiratory pressure (Pmean), plateau inspiratory pressure (Pplateau), and expired Vt were recorded at the end of the following study periods: (1) during two-lung ventilation using VCV before initiation of OLV (baseline); (2) during OLV 30 min after the first randomized ventilation mode; (3) 30 min after the second ventilation mode; and (4) 20 min after reestablishing two-lung ventilation after pulmonary resection. There was no surgical manipulation of the nondependent lung during measurements and the first three measurements were recorded before any major pulmonary vessel clipping.
To detect a difference of 40 mm Hg in Pao2 during OLV with a two-sided approximation accepting an α error of 5% and a β error of 20%, the required study size was calculated to be 52 patients. To account for patients dropping out during the study, 10% more patients were added, for a final sample size of 58 patients. All data were expressed as mean and standard deviation (sd). Statistical analysis was performed using the two-way ANOVA for repeated measures and a paired t-test. A P < 0.05 was considered statistically significant.
One patient in group A had to be withdrawn from the study due to atrial fibrillation with severe hypotension. Fifty-seven patients completed the study. Demographic characteristics, preoperative lung function, baseline Pao2, Paco2, and the surgical procedures performed are listed in Table 1. There were no statistically significant differences in demographic or clinical data between groups (Table 1), although left thoracotomies were more common in group B (P = 0.024). The majority of studied patients had normal preoperative pulmonary function; a minority of patients (37%) had moderate or severe obstructive and restrictive respiratory pattern (Table 2). The distribution of these patients was similar between intervention groups.
Data obtained during OLV with VCV or PCV are listed in Table 3. There were no differences in Pao2 between OLV with VCV compared with PCV (P = 0.534). Compared with two-lung ventilation, initiation of OLV with either VCV or PCV was associated with a significant increase in peak (P < 0.001), plateau (P < 0.001), and mean (P < 0.001) airway pressures while Pao2 decreased (P < 0.001). Peak airway pressures were higher (P < 0.001) during OLV with VCV than during PCV. There were no significant differences in Pao2 between OLV with VCV compared with PCV (P = 0.534). The sequence of randomization had no influence on Pao2 (P = 0.353). There were no episodes of Sao2 <90% in either group. There were no significant differences in Paco2, ETco2 or mean airway pressure during OLV with either VCV or PCV. In patients with obstructive respiratory disease (the percent of predicted forced expiratory volume in 1 s (FEV)1 62.4 ± 9.1 and FEV1/forced vital capacity (FVC) 60%) there was no difference in Pao2 between OLV-VCV versus OLV-PCV (VCV, 225.00 ± 66.3 mm Hg vs. PCV, 213.54 ± 62.1 mm Hg, P = 0.341). In patients with restrictive respiratory disease (the percent of predicted FVC 56.4 ± 16.7 and FEV1/FVC 87%) OLV with VCV versus PCV did not result in differences in Pao2 (OLV-VCV, 256.60 ± 63.77 mm Hg vs. PCV, 210.80 ± 61.86 mm Hg, P = 0.172).
Using the nomogram of Benatar et al., (10) we found that during two-lung ventilation the mean intrapulmonary shunt ranged from 7% to 22% (mean 15%). During OLV with VCV the shunt fraction ranged from 16% to 40% (mean 24%), and during OLV with PCV the shunt was from 16% to 42% (mean 24%). Therefore, the shunt during OLV was similar with VCV or PCV.
PCV is suggested as a rational method of ventilation during OLV to ensure oxygenation while minimizing peak airway pressure (11). The results of this study, however, were that PCV provides no benefit for arterial oxygenation during OLV compared to VCV. Peak airway pressures were predictably lower with PCV than VCV.
Our results are consistent with those of previous studies comparing VCV to PCV in patients with ALI or ARDS, which found that PCV offers no advantage over VCV for improving oxygenation (5–7). More recently, Prella et al. (4) studied the effects of PCV on gas exchange and gas distribution using computerized axial tomography scan quantitative analysis of lung densities in patients with ALI/ARDS. These investigators found that Pao2 and the surface area of the nonaerated lung zones were unchanged with PCV compared with VCV (4). In that study, VCV did result in a larger nonaerated area in the apex of the lung compared with PCV.
Our results contrast with those of Tugrul et al. (2) who, in a cross-over trial of 48 thoracic surgical patients, found that PCV resulted in improved oxygenation during OLV compared with VCV. The higher pulmonary shunt they observed during VCV compared with PCV was attributed to higher plateau pressure with this mode of ventilation. They further found that patients whose Pao2 improved with PCV generally had a lower FVC. They suggested that patients with restrictive pulmonary disease might benefit from PCV, even though there was only a weak correlation between Pao2 values and FVC (r = −0.3). The number of patients with a restrictive respiratory pattern in our study was too small to assess the effects of ventilatory mode on arterial oxygenation in this subset of patients. In our study left thoracotomies predominated, whereas in Tugrul et al.'s study right thoracotomies were more frequent (2). Pao2 during OLV to the right lung (i.e. left thoracotomy) is higher than during OLV to the left lung (i.e. during right thoracotomy) owing to the larger lung area, and thus less intrapulmonary shunt (12). Therefore, our results might be attributed to this larger proportion of left-sided thoracic surgical procedures. In our study, though, each patient served as their own control, sequentially undergoing both ventilatory modes. Thus, differences in the surgical site compared with Tugrul et al.'s study (2) cannot necessarily explain our findings. Another explanation for the divergent findings was the overall better pulmonary function before surgery in our series compared with the patients in the study by Tugrul et al. (2). Of note, the Paco2 and ETco2 were not significantly different in OLV-VCV or OLV-PCV, suggesting the same efficiency of ventilation of both ventilatory techniques.
The main advantage of PCV versus VCV appears to be lower peak airway pressure that might decrease the risk for barotrauma during mechanical ventilation (2,4,6). Nevertheless, in a prospective randomized trial comparing VCV to PCV in 79 patients with ARDS, there were no statistically significant differences in the incidence of barotrauma between ventilatory modes (6). Peak airway pressure does not reflect peak alveolar pressure: peak airway pressure is much greater, and depends on endotracheal tube resistance, inspiratory flow, and the respiratory mechanics of the lung (13). Further, there appears to be only a weak correlation between peak airway pressure and the incidence of barotrauma (14). In contrast, there is a strong correlation between plateau airway pressure and mechanical ventilation-induced barotrauma when plateau airway pressure levels exceed 35 cm H2O (14). Therefore, during VCV the risk of lung injury can be minimized by limiting the plateau airway pressure to <30 cm H2O (15). Other data suggest that, in ventilator-induced lung injury in rabbits, high peak inspiratory flow with PCV induces significantly more severe lung damage manifested by hypoxemia and histological damage than lower peak inspiratory flows determined by VCV (9). Finally, in the setting of rapid changes in lung impedance, PCV may not be able to deliver a constant Vt, or it may result in autopositive end expiratory pressure in patients with obstructive lung disease leading to altered hemodynamics and reduced Vt (16,17).
In conclusion, primarily in patients with normal preoperative lung function we found that the ventilatory mode during OLV had little influence on gas exchange.
The authors thank Mr. Francesc Fontana, bioengineering technician, for his help and technical assistance.
1. Slinger P. Management of one-lung anesthesia. IARS 2005 review course lectures. Anesth Analg 2005;Suppl:89–94.
2. Tugrul M, Çamci E, Karadeniz H, et al. Comparison of volume controlled with pressure controlled ventilation during one-lung anaesthesia. Br J Anaesth 1997;79:306–10.
3. MacIntyre NR. New modes of mechanical ventilation. Clin Chest Med 1996;17:411–21.
4. Prella M, Feihl F, Domenighetti G. Effects of short-term pressure-controlled ventilation on gas exchange, airway pressures, and gas distribution in patients with acute lung injury/ARDS: comparison with volume-controlled ventilation. Chest 2002;122:1382–8.
5. Rappaport SH, Shpiner R, Yoshihara G, et al. Randomized, prospective trial of pressure-limited versus volume-controlled ventilation in severe respiratory failure. Crit Care Med 1994;22:22–32.
6. Esteban A, Alía I, Gordo F, et al. Prospective randomized trial comparing pressure-controlled ventilation and volume-controlled ventilation in ARDS. Chest 2000;117:1690–6.
7. Edibam C, Rutten A, Collins D, Bersten A. Effect of inspiratory flow pattern and inspiratory to expiratory ratio on nonlinear elastic behaviour in patients with acute lung injury. Am J Respir Crit Care Med 2003;167:702–7.
8. Dembinski R, Henzler D, Bensberg R, et al. Ventilation-perfusion distribution related to different inspiratory flow pat-terns in experimental lung injury. Anesth Analg 2004;98:211–19.
9. Maeda Y, Fujino Y, Uchiyama A, et al. Effects of peak inspiratory flow on development of ventilator-induced lung injury in rabbits. Anesthesiology 2004;101:722–8.
10. Benatar SR, Hewlett AM, Nunn JF. The use of iso-shunt lines for control of oxygen therapy. Br J Anaesth 1973;45:711–18.
11. Sentürk M. New concepts of the management of one-lung ventilation. Curr Opin Anaesthesiol 2006;19:1–4.
12. Slinger P, Johnston M. Preoperative evaluation of the thoracic surgery patient. In: Kaplan J, Slinger P, eds. Thoracic anaesthesia. Philadelphia: Churchill Livingstone, 2003:1–23.
13. Kawati R, Lattuada M, Sjöstrand U, et al. Peak airway pressure increase is a late warning sign of partial endotracheal obstruction whereas change in expiratory flow is an early warning sign. Anesth Analg 2005;100:889–93.
14. Boussarsar M, Thierry G, Jaber S, et al. Relationship between ventilatory settings and barotrauma in the acute respiratory distress syndrome. Intensive Care Med 2002;28:406–13.
15. MacIntyre N. Setting the frequence—-tidal volume pattern. Respir Care 2002;47:266–74.
16. Campbell R, Davis B. Pressure-controlled versus volume-controlled ventilation: does it matter? Respir Care 2002;47:416–24.
17. Ducros L, Moutafis M, Castelain MH, et al. Pulmonary air trapping during two-lung and one-lung ventilation. J Cardiothorac Vasc Anesth 1999;13:35–9.