This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (2) Citing Articles via Google Scholar Google Scholar Articles by Szegedi, L. L. Articles by d’Hollander, A. A. Search for Related Content PubMed PubMed Citation Articles by Szegedi, L. L. Articles by d’Hollander, A. A. Related Collections Cardiovascular Blood Anesthetic Techniques Monitoring (Non-cardiac)

---

CARDIOVASCULAR ANESTHESIA

The Effects of Acute Isovolemic Hemodilution on Oxygenation During One-Lung Ventilation

Laszlo L. Szegedi, MD^*, Philippe Van der Linden, MD, PhD, Anne Ducart, MD, Pieter Cosaert, MD^*, Jan Poelaert, MD, PhD, Frank Vermassen, MD, PhD^||, Eric P. Mortier, MD, DSc^*, and Alain A. d’Hollander, MD, PhD^¶

*Department of Anesthesiology, Gent University Hospital, Gent, Belgium; Department of Anesthesiology, Brugmann University Hospital, Brussels, Belgium; Department of Anesthesiology, Erasme University Hospital, Brussels, Belgium; Department of Cardiac Anesthesia and Intensive Care, Gent University Hospital, Gent, Belgium; ||Department of Thoracic and Vascular Surgery, Gent University Hospital, Gent, Belgium; and ¶Department of Anesthesiology, Geneva University Hospital, Geneva, Switzerland

Address correspondence and reprint requests to Laszlo L. Szegedi, MD, Department of Anesthesiology 2K12, Gent University Hospital, De Pintelaan, 185, 9000 Gent, Belgium. Address e-mail to laszlo.szegedi{at}ugent.be

	Abstract

Top Abstract Introduction Methods Results Discussion References

 
Data on the effects of isovolemic hemodilution (IH) on oxygenation during one-lung ventilation (OLV) are lacking. We studied 47 patients with hemoglobin >14 g/dL who were scheduled for lung surgery (17 with normal lung function [group NL], 17 with chronic obstructive pulmonary disease [COPD] [group COPD], and 13 with COPD as control for time/anesthesia effects [group CTRL]). Anesthesia was standardized. The tracheas were intubated with a double-lumen tube. Ventilatory settings and fraction of inspired oxygen remained constant. The study was performed with patients in the supine position before surgery. OLV was initiated for 15 min. Two-lung ventilation was reinstituted, and IH was performed (500 mL); an identical volume of hydroxyethyl starch was administered. Subsequently, OLV was again performed for 15 min. In group CTRL, the same sequences of OLV were performed without IH. At the end of each period of OLV, pulmonary mechanics and blood gases were recorded. Data were analyzed by analysis of variance (mean ± SD). In group NL and group CTRL, the arterial oxygen partial pressure remained constant, whereas it decreased in group COPD from 119 ± 21 mm Hg before IH to 86 ± 16 mm Hg after IH (P < 0.01). Mild IH impairs gas exchange during OLV in COPD patients, but not in patients with normal lung function.

IMPLICATIONS: This clinical study suggests that hemodilution impairs gas exchange during one-lung ventilation in patients with pulmonary hyperinflation and chronic obstructive lung disease, but not in patients with normal lungs; the reasons for this finding are not clear.

	Introduction

Top Abstract Introduction Methods Results Discussion References

 
One-lung ventilation (OLV) is a useful technique during thoracic surgery (1). One lung is mechanically ventilated while the other is either occluded or left open to the atmosphere. The challenge for the anesthesiologist during OLV is to maintain acceptable oxygenation. Indeed, blood flow through the nonventilated lung becomes a right-to-left shunt in addition to that which exists in the ventilated lung. OLV will therefore result in a lower arterial oxygen partial pressure (PaO₂) than during two-lung ventilation (TLV). Fortunately, several mechanisms tend to decrease the percentage of cardiac output (CO) perfusing the nonventilated lung. These mechanisms are either passive (e.g., mechanical-like gravity, surgical manipulation, amount of preexisting lung disease) or active (e.g., hypoxic pulmonary vasoconstriction [HPV]) (1).

Increasing awareness of potential severe complications associated with blood transfusion has prompted the development of blood-conservation strategies. Among these strategies, acute isovolemic hemodilution (IH) is a safe procedure (2). The IH technique decreases a patient’s hemoglobin (Hb) at the start of the operation so fresh units of the patient’s blood are available when needed (2).

Few studies have investigated the effects of hemodilution on pulmonary gas exchange, particularly during OLV. Acute hemodilution may improve the gas-exchange efficiency of the normal lung (3,4), as evidenced by higher PaO₂ and lower alveolar-arterial oxygen partial pressure difference, but this effect is not consistently observed. Experimentally, the relationship between anemia and PaO₂ is not well defined. Deem et al. (4) analyzed studies concerning the relationship between anemia and PaO₂. These studies have examined the effects of hemodilution or blood transfusion in the absence of known lung disease, suggesting an inverse relationship between PaO₂ and hematocrit (Hct), a direct relationship, or an inconsistent effect. Moreover, only a few studies included concurrent controls for the effects of time, anesthesia, or both (4).

The purpose of this study was to assess the effect of IH on arterial oxygenation during OLV. Patients scheduled for thoracic surgery may have compromised preoperative lung function. Therefore, we evaluated the effect of IH on arterial oxygenation in patients with normal preoperative lung function and in patients with stable chronic obstructive pulmonary disease (COPD) with moderate degrees of pulmonary hyperinflation. A third group of patients with COPD who did not undergo IH was added to serve as control for the effects of time, anesthesia, or both.

	Methods

Top Abstract Introduction Methods Results Discussion References

 
Our IRB approved the investigation, and informed consent was obtained from 47 adult ASA class II–III patients with an Hb level >140 g/L. The patients were scheduled for tumor surgery through right thoracotomy and required OLV. Preoperative pulmonary function—including spirometry and static lung volumes determined by plethysmography (MasterScreen Body^TM, Jaeger and Toennies^TM; Erich Jaeger GmbH, Hoechberg, Germany)—was tested with patients in the sitting position.

On the basis of their preoperative pulmonary function, patients were included either in a group with normal lung function (group NL; n = 17) or in a group with stable COPD and mild to moderate pulmonary hyperinflation (functional residual capacity more than 120% of the predicted value) (group COPD; n = 17). After preliminary results on 12 patients in each group, a third group of patients with stable COPD was added (group CTRL; n = 13). The assignment of the patients either to group COPD or to group CTRL was randomized. The rationale for including a concurrent control group (group CTRL) was to eliminate the effects of time, anesthesia, or both that could influence our results. Hence, these patients underwent the same sequences of OLV but no IH. Patients with cardiac or renal function impairment were not included in the study.

Anesthesia was conducted in a standardized manner. All patients were given oral alprazolam 0.5 mg approximately 60 to 90 min before arrival in the operating room.

A thoracic epidural catheter was inserted at the midthoracic level (T6 to T9) to ensure analgesia during and after surgery. A test dose of 3 mL of 2% lidocaine with epinephrine 1:200,000 was used; the initial dose (8 mL of 2% lidocaine with 100 µg of fentanyl) was given only after the end of the study, followed by a continuous infusion of bupivacaine 0.5% at 2–5 mL/h.

The induction and maintenance of general anesthesia was achieved with fentanyl (100 µg) and propofol (initial dose of 2 mg/kg followed by continuous infusion of 3–5 mg · kg^–1 · h^–1). Cisatracurium (initial dose of 0.15 mg/kg and top-up doses of 0.05 mg/kg) was used to allow tracheal intubation and to maintain neuromuscular blockade throughout surgery. The neuromuscular blockade was assessed by regular measurements of posttetanic count during the procedure.

A three-lead electrocardiogram, invasive radial arterial and central venous blood pressures, and arterial oxygen saturation were continuously monitored. Expired end-tidal carbon dioxide tension and flow-volume and pressure-volume loops were also continuously displayed (S/5 Anesthesia Monitor AM; Datex-Ohmeda Division, Instrumentarium Corp., Datex-Ohmeda, Finland).

In all patients, the bronchus of the dependent left lung was intubated with a double-lumen endotracheal tube (DLT) (Broncho-cath^TM; Mallinckrodt Laboratories, Athlone, Ireland) of an appropriate size determined by the height of the patient (5). The correct position of the DLT was ascertained with fiberoptic bronchoscopy.

A constant tidal volume of 10 mL/kg was delivered with an S/5 ADU ventilator throughout the study. The ventilatory pattern consisted of a volume-controlled square-wave flow pattern at a rate of 10 breaths/min, with a fraction of inspired oxygen in air of 0.5. Inspiratory time/expiratory time was 1:2, and the end-inspiratory pause was 10% of the total respiratory cycle. End-expiratory pressure was set to 0. Ventilatory variables were kept constant during the study during both TLV and OLV.

A transesophageal echocardiograph probe was inserted after tracheal intubation to measure the CO with the method of the effective aortic valve area (6). Blood gas samples were analyzed immediately after they were drawn and were temperature-corrected. Arterial and central venous oxygen saturation were measured with a cooximeter.

This investigation was performed with closed chest before the surgical procedure, with the patients in the supine position. After intubation in the supine position and fiberoptic bronchoscopic control of the correct DLT position, the tracheal lumen of the DLT was clamped, and the nonventilated right lung was allowed to deflate to atmospheric pressure. After 15 min of OLV with the above-described ventilatory settings, end-inspiratory and end-expiratory occlusions were performed to determine the mechanical characteristics of the respiratory system (peak inspiratory airway pressure, end-inspiratory plateau pressure, and intrinsic positive end-expiratory pressure [PEEP_i]), and arterial and venous blood gas samples were drawn and analyzed. The CO was measured.

After these baseline measurements (before hemodilution), the lung was sighed manually (insufflation pressure up to 40 cm H₂O) (7), and TLV was restored with unaltered ventilatory settings. An IH was performed during the period of TLV by simultaneous withdrawal of blood and infusion of 6% hydroxyethyl starch 130/0.4 at equal volumes. The exchange was standardized to 500 mL for each patient.

Thereafter, OLV was restored for 15 min. At the end of this period, ventilatory data were again recorded, arterial and venous gas samples were drawn and analyzed, and CO was measured (after IH).

In group CTRL, the sequences of OLV were the same but no IH was performed. The temperature of the patients was kept constant during the entire study with an air convection system.

Statistical analysis was performed with the GraphPad InStat software package, Version 3.05, for Windows 95/NT. First, the assumption that data were sampled from populations with identical standard deviations was tested by using the method of Bartlett. The assumption that the differences were sampled from populations that follow Gaussian distribution was verified by using the method of Kolmogorov and Smirnov. Thereafter, differences between groups and inside each group were analyzed with one-way analysis of variance (ANOVA) with the Tukey-Kramer multiple comparisons test. Values of P < 0.05 were accepted as statistically significant. Data are presented as mean ± SD.

	Results

Top Abstract Introduction Methods Results Discussion References

 
There was a slight, but statistically significant, difference when comparing the weight of the patients from group NL with that of those from either group COPD or group CTRL, but there were no differences in body-surface area (Table 1). Forced expired volumes in 1 s were significantly smaller, and functional residual capacity and residual volumes were significantly larger in group COPD and in group CTRL as compared with group NL (Table 1).

View larger version (19K): [in this window] [in a new window]   Figure 1. Individual values of the partial pressure of oxygen in arterial blood (PaO2) in the three groups of patients at the two time periods studied. The PaO2 decreased significantly after isovolemic hemodilution in group COPD (n = 17), but no decreases were found in group NL (n = 17) or group CTRL (n = 13) (one-way analysis of variance [ANOVA] with the Tukey-Kramer multiple comparisons test). Group NL = group of patients with normal lung function; group COPD = group of patients with chronic obstructive pulmonary disease; group CTRL = control group (patients with chronic obstructive pulmonary disease who had the same sequences of one-lung ventilation but who did not undergo hemodilution); IH = isovolemic hemodilution; OLV 1 = first period of one-lung ventilation in group CTRL; OLV 2 = second period of one-lung ventilation in group CTRL; NS = not significant.

	Discussion

Top Abstract Introduction Methods Results Discussion References

HPV is the major protective mechanism that diverts blood flow away from the atelectatic lung (1). Although HPV is an intrinsic mechanism of the pulmonary vasculature (8), several studies indicate that it may be influenced by red blood cells. McMurtry et al. (9) reported that isolated rat lungs perfused with plasma had rapidly decaying HPV when compared with lungs perfused with blood. Another study reported similar results in isolated rat, cat, or rabbit lungs (10).

Deem et al. (11) studied the effect of anemia on intrapulmonary shunt during left lung atelectasis in rabbits and concluded that isovolemic anemia has a deleterious effect on pulmonary gas exchange, possibly through attenuation of HPV. Another study (12) suggested that in rabbits with previous lung injury induced by gas embolism, IH resulted in improved oxygen exchange, but it did not find a clear mechanism for this improvement. The same authors reported an improvement in gas exchange in the normal rabbit lung as a result of an improvement in overall ventilation/perfusion matching (13). Kleen et al. (14) suggested that severe acute normovolemic hemodilution in dogs causes different alterations in the heterogeneity of regional pulmonary blood flow in hyperoxic conditions. Other authors (15) found no influence of changes in Hb concentration on HPV efficiency.

In this study, preexisting lung disease, as demonstrated by altered preoperative pulmonary function tests, was associated with a higher baseline PEEP_i and a slightly lower PaO₂ (group COPD and group CTRL). The most important finding of this study was a significant decrease in PaO₂ after IH during OLV of the patients with compromised lung function. COPD patients are chronically hypoxic; thus, in such patients HPV may be already maximal and no more protective during OLV, because these patients already have an increased pulmonary vascular resistance and decreased pulmonary vascular bed (16). Other mechanisms that tend to decrease the degree of venous admixture during OLV include gravitational effect and surgical manipulation (1). In this study, performed with a closed chest and with patients in the supine position, these potential factors were absent.

The maintenance of tissue oxygen delivery during an acute reduction in red blood cell concentration depends on both an increase in CO and an increase in blood oxygen extraction (17). However, in our study, CO remained constant with the mild IH (Table 2). Anesthesia significantly reduces the CO response associated with acute IH. This could be related to the effects of the anesthetics on the autonomic and the cardiovascular systems (18).

The higher PEEP_i values (Table 2) during OLV observed in group COPD as compared with group NL could have induced enough hyperinflation to disrupt the ventilation/perfusion matching. However, this observation is unlikely, given that in group CTRL (COPD patients who did not undergo IH) the PEEP_i values were comparable to those measured in group COPD and that the PaO₂ changes in group CTRL were not significant.

If the PaO₂ was lower in group COPD after IH than in group CTRL (without IH) and if there were no differences in the oxygen saturation of blood collected in the superior cava vein or in the cardiac index, then the shunt must have increased in group COPD; even though difficult to explain, this may be the only reason for our findings.

Some consideration should be given to the limitations of our study. First, it was a purely clinical study; according to institutional structures, monitoring and measurements were limited. For example, a pulmonary artery catheter could not be inserted, so we were not able to perform shunt measurements. Second, given the absence of data on the subjects and the patient population studied (stable COPD with mild to moderate pulmonary hyperinflation), only a mild degree of IH was studied. Further investigations are required to evaluate the effects of more profound IH in patients with normal preoperative lung function.

It is surprising that this mild hemodilution/anemia was associated with this degree of gas-exchange impairment in the patients with COPD, given that changes in intrapulmonary shunt in the experimental model reported by Deem et al. (11) were not realized until the Hct was in the mid-20s. The different results in the experimental model and this clinical study could be for any number of reasons, including species differences and the presence of COPD, but it raises the question as to whether the observed decrease in PaO₂ was in fact due to hemodilution. Although this is somewhat unlikely (18), the study protocol allows the possibility that the observed change in arterial oxygenation occurred after IH during TLV.

Third, our results could have been influenced by the repeated endotracheal tube clamping maneuvers. Benumof (19) showed that intermittent hypoxic challenges may potentiate the HPV response in an animal model. In our study design, there were, indeed, two sequences of deflating/inflating of the lung; however, the oxygenation was not better in either group in the second period of OLV. This is in accordance with the studies of Carlsson et al. (20), who found maximal HPV within 15 minutes of hypoxia, and Domino et al. (21), who observed a maximal response after the first hypoxic challenge during OLV in closed-chest dogs. To eliminate the changes over time in oxygenation during OLV, a third group of patients (group CTRL) with altered preoperative pulmonary function was studied. Sequences of OLV without IH did not result in significant difference in arterial oxygenation (Table 3).

Fourth, the study was performed with the patients in the supine position and with a closed chest instead of the usual clinical situation of a lateral position with an open chest. These differences may affect extrapolation of the results because of the absence of the gravitational effect in COPD patients and the absence of surgical manipulation—factors that may contribute to blood-flow redistribution during OLV (1).

In conclusion, mild IH significantly altered arterial oxygenation during OLV in patients with preoperative compromised lung function. Although anemia may be less tolerated by COPD patients in these conditions, this does not indicate that a more aggressive transfusion approach is required in these patients. Indeed, PaO₂ could be easily increased by increasing the fraction of inspired oxygen, which is common practice during OLV. Acute IH did not appear to be a safe alternative technique to blood transfusion in patients with altered pulmonary function undergoing thoracic surgery. Further studies are required to better define the adequate transfusion trigger in this population.

	References

Top Abstract Introduction Methods Results Discussion References

 

1. Benumof JL. Special respiratory physiology of the lateral decubitus position, the open chest and one-lung ventilation. In: Benumof JL, ed. Anesthesia for thoracic surgery. Philadelphia: WB Saunders, 1995: 123–51.
2. Goodnough LT, Brecher ME, Kanter MH, AuBuchon JP. Transfusion medicine: first of two parts—blood transfusion. N Engl J Med 1999; 340: 438–47.[Free Full Text]
3. Deem S, Alberts MK, Bishop MJ, et al. CO₂ transport in normovolemic anemia: complete compensation and stability of blood CO₂ tensions. J Appl Physiol 1997; 83: 1240–6.
4. Deem S, Swenson ER, Alberts MK, et al. Red-blood-cell augmentation of hypoxic pulmonary vasoconstriction: hematocrit dependence and the importance of nitric oxide. Am J Respir Crit Care Med 1998; 157: 1181–6.[Abstract/Free Full Text]
5. Hannallah MS, Benumof JL, McCarthy PO, et al. Comparison of three techniques to inflate the bronchial cuff of left polyvinyl chloride double-lumen tubes. Anesth Analg 1993; 77: 990–4.[Abstract/Free Full Text]
6. Poelaert J, Schmidt C, Van Aken H, et al. A comparison of transoesophageal echocardiographic Doppler across the aortic valve and thermodilution technique for estimating cardiac output. Anaesthesia 1999; 54: 128–36.[Web of Science][Medline]
7. Rothen HU, Sporre B, Englberg G, et al. Re-expansion of atelectasis during general anaesthesia: a computed tomography study. Br J Anaesth 1993; 71: 788–95.[Abstract/Free Full Text]
8. Voelkel NF. Mechanism of hypoxic pulmonary vasoconstriction. Am Rev Respir Dis 1986; 133: 1186–95.[Web of Science][Medline]
9. McMurtry IF, Hookway BW, Roos SD. Red blood cells play a crucial role in maintaining vascular reactivity to hypoxia in isolated rat lungs. Chest 1977; 71: 253–6.[Free Full Text]
10. Hakim TS, Malik AB. Hypoxic vasoconstriction in blood and plasma perfused lungs. Respir Physiol 1988; 72: 109–21.[Web of Science][Medline]
11. Deem S, Bishop MJ, Alberts MK. Effect of anemia on intrapulmonary shunt during atelectasis in rabbits. J Appl Physiol 1995; 79: 1951–7.[Abstract/Free Full Text]
12. Deem S, McKinney S, Polissar NL, et al. Hemodilution during venous gas embolisation improves gas exchange, without altering V(A)/Q or pulmonary blood flow distributions. Anesthesiology 1999; 91: 1861–72.[Web of Science][Medline]
13. Deem S, Hedges RG, McKinney S, et al. Mechanism of improvement in pulmonary gas exchange during isovolemic hemodilution. J Appl Physiol 1999; 87: 132–41.[Abstract/Free Full Text]
14. Kleen M, Habler O, Hutter J, et al. Hemodilution and hyperoxia locally change distribution of regional pulmonary perfusion in dogs. Am J Physiol Heart Circ Physiol 1995; 274: H520–8.
15. Brimioulle S, Lejeune P, Naeije R. Effects of hypoxic pulmonary vasoconstriction on pulmonary gas exchange. J Appl Physiol 1996; 81: 1535–43.[Abstract/Free Full Text]
16. Slinger P. New trends in anaesthesia for thoracic surgery including thoracoscopy. Can J Anaesth 1995; 42: R77–84.
17. Van der Linden P. Anemic hypoxia. In: Sibbald WJ, Messmer K, Fink MP, eds. Tissue oxygenation in acute medicine. Berlin: Springer-Verlag, 1998: 116–27.
18. Ickx BE, Rigolet M, Van Der Linden PJ. Cardiovascular and metabolic response to acute normovolemic anemia: effects of anesthesia. Anesthesiology 2000; 93: 1011–6.[Web of Science][Medline]
19. Benumof JL. Intermittent hypoxia increases lobar hypoxic pulmonary vasoconstriction. Anesthesiology 1983; 58: 399–404.[Web of Science][Medline]
20. Carlsson AJ, Bindslev L, Santesson J, et al. Hypoxic pulmonary vasoconstriction in the human lung: the effect of prolonged unilateral hypoxic challenge during anaesthesia. Acta Anaesthesiol Scand 1985; 29: 346–51.[Web of Science][Medline]
21. Domino K, Chen L, Alexander C, et al. Time course and responses of sustained hypoxic pulmonary vasoconstriction in the dog. Anesthesiology 1984; 60: 562–6.[Web of Science][Medline]

This article has been cited by other articles:

				A. I. Levin and J. F. Coetzee Arterial Oxygenation During One-Lung Anesthesia Anesth. Analg., January 1, 2005; 100(1): 12 - 14. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (2) Citing Articles via Google Scholar Google Scholar Articles by Szegedi, L. L. Articles by d’Hollander, A. A. Search for Related Content PubMed PubMed Citation Articles by Szegedi, L. L. Articles by d’Hollander, A. A. Related Collections Cardiovascular Blood Anesthetic Techniques Monitoring (Non-cardiac)