The amount of oxygen consumed by the lung itself is difficult to measure because it is included in whole-body gas exchange. It may be increased markedly under pathological conditions such as lung infection or adult respiratory distress syndrome. To estimate normal oxygen consumption of the human lung as a basis for further studies, respiratory gas analysis during total cardiopulmonary bypass may be a simple approach because the pulmonary circulation is separated from systemic blood flow during this period.


Lung oxygen consumption was determined in 16 patients undergoing cardiac surgery. During total cardiopulmonary bypass their lungs were ventilated with low minute volumes (tidal volume, 150 ml; rate, 6 min-1; inspiratory oxygen fraction, 0.5; positive end-expiratory pressure, 3 mmHg). All expiratory gas was collected and analyzed by indirect calorimetry. As a reference value also, whole-body oxygen consumption of these patients was determined before total cardiopulmonary bypass. In a pilot study of eight additional patients (same ventilatory pattern), the contribution of systemic (bronchial) blood flow to pulmonary gas exchange during cardiopulmonary bypass was assessed. For this purpose, the amount of enflurance diffusing from the systemic blood into the bronchial system was measured.


The human lung consumes about 5-6 ml oxygen per minute at an esophageal temperature of 28 degrees C. Prebypass whole-body oxygen consumption measured at nearly normothermic conditions was 198 +/- 28 ml/min. Mean lung and whole-body respiratory quotients were similar (0.84 and 0.77, respectively). Extrapolating lung oxygen consumption to 36 degrees C suggests that the lung consumes about 11 ml/min or about 5% of total body oxygen consumption. Because the amount of enflurane diffused from the systemic circulation into the bronchial system during cardiopulmonary bypass was less than 0.1%, the contribution of bronchial blood flow to lung gas exchange can be assumed to be negligible.


The lung consumes about 5% of whole-body oxygen uptake.

The lung fulfills specialized and energy-consuming functions including tracheobronchial clearance, regulation of distribution of air and blood flow, and surfactant turnover. [1,2]Therefore, the constituent cells of the lung have metabolic requirements that must be satisfied to maintain functional and structural integrity. This amount may be increased markedly under pathologic conditions such as lung infection or adult respiratory distress syndrome.

Because lung-specific gas exchange is incorporated in the gas exchange of the whole body, it is difficult to gain precise information about oxygen uptake and carbon dioxide excretion of the lung tissue itself as two major variables of pulmonary metabolism. Therefore, most data related to lung tissue oxygen consumption have been provided by indirect approaches using lung slices [3–5]or isolated lung preparations. [6–9]In the intact organism, the difference between whole-body oxygen consumption determined by indirect calorimetry and the Fick principle has been used to estimate the oxygen consumption of the lung, [10,11]because the latter measures only systemic arteriovenous oxygen content difference, excluding the lung. All three approaches have clear limitations permitting only rough estimations of the lung's true oxygen consumption.

For direct measurements of lung oxygen consumption in humans, the lungs have to be separated from pulmonary arterial blood flow. This situation is realized during extracorporeal circulation for cardiac operations when, during total cardiopulmonary bypass (CPB), all blood from the superior and inferior vena cava is drained into a reservoir from which a pump delivers oxygenated blood into the ascending aorta. During this period it becomes possible to determine oxygen uptake and carbon dioxide excretion of the lung itself by respiratory gas analysis. Thus, this approach yields direct measurements of true lung oxygen consumption in humans.


Sixteen patients (see Table 1for demographic data) scheduled for open heart surgery (coronary bypass grafting or valvular replacement) were included in this study. With the exception of two patients with pulmonary arterial hypertension due to mitral valve regurgitation, no patient had any further relevant pulmonary disorders. Specifically, no patient suffered from pulmonary edema, pneumonia, bronchial asthma, sepsis, adult respiratory distress syndrome, or had pathological respiratory lung function tests before operation (vital capacity, forced expired minute volume, blood gas analysis). The investigation was approved by the institutional ethical committee, and informed consent was obtained from each patient.

Anesthesia and Cardiopulmonary Bypass

After premedication (10 mg diazepam given orally, 25 mg meperidine, 25 mg promethazine, and 1.25 mg droperidol given intramuscularly) and induction of anesthesia (1–2 mg/kg thiopental, 1–3 micro gram/kg fentanyl, and 0.1 mg/kg pancuronium), the lungs were ventilated mechanically maintaining normocapnia (Servo 900 D; inspiratory oxygen fraction, 0.5%; respiratory rate, 10 breaths/min). Anesthesia was maintained by ventilation with enflurane (0.4 to 1% inspired) supplemented with additional doses of fentanyl (0.1 to 0.2 mg) before noxious stimuli such as skin incision or sternotomy.

Cardiac arrest and hypothermia of the myocardium was achieved by intracoronary infusion of cardioplegic solution (modified Bretschneider) in association with systemic hypothermia (28 degrees Celsius). During total CPB, all blood from the caval veins was drained into the cardiotomy reservoir and the caval veins were occluded by banding so that pulmonary artery flow ceased completely. The left ventricle was vented in all cases. Blood was oxygenated by a membrane oxygenator (Cobe, Arvada, CO), and a nonpulsatile flow rate of 2–3 l [center dot] min sup -1 [center dot] m sup -2 was regulated by a roller pump (Stockert, Munich, Germany) to a mean arterial pressure of 50–80 mmHg. Systemic hypothermia (28 degrees Celsius, alpha-stat pH management) was induced by blood cooling and maintained for at least 30 min. During this period, the lungs were ventilated with low tidal volumes (150 ml) at a rate of 6 breaths/min and with a positive end-expiratory pressure of 3 mmHg. Nasopharyngeal, rectal, and blood temperatures (during CBP) were monitored continuously.


Lung oxygen uptake (V with dot O2), carbon dioxide excretion (V with dot CO2), and the respiratory quotient were determined in 1-min intervals with an indirect calorimeter (Deltatrac II, Datex Instrumentarium, Helsinki, Finland). All expired gas was completely collected and analyzed using the baby mode of the calorimeter (diluting flow in the mixing chamber, 3 l/min). The relative error using this mode of the metabolic monitor had been validated before (+/- 1.5–2% in the measurement of V with dot CO2and 1.9–4% for V with dot O2). [12,13]In our study, the average coefficient of variation of lung V with dot2determinations during the minute-to-minute registrations of the study period was 7.8 +/- 1.6%.

The calorimeter was calibrated before each study with a high-accuracy calibration gas (mixture of 95% oxygen and 5% carbon dioxide). To ensure that all expired gas was analyzed, the breathing circuit was checked meticulously and the pressure of the tracheal cuff was controlled to avoid gas leakage.

In addition to lung V with dot2, whole-body V with dot O sub 2 at about 36 degrees Celsius (esophageal temperature) was determined before CPB with the adult mode of the calorimeter (diluting flow, 40 l/min) during a period with stable hemodynamic conditions (no inotropic support).

To assess the contribution of the bronchial circulation to pulmonary gas exchange during total CPB, we measured in a pilot study (eight additional patients, same ventilatory pattern during total CPB: tidal volume, 150 ml; rate, 6 breaths/min) the amount of enflurane transferred from the systemic circulation to the alveolar space. For this purpose the gas stream to the oxygenator (50% oxygen in air; flow 2.5 l/min) was conducted through a Drager vaporizer (Lubeck, Germany) and contained 2% enflurane. A specially designed probe was introduced via the tracheal tube to a final position 2 cm above the carina. From there, bronchial gas was sampled (suction flow, 50 ml/min; size of sampling chamber, 4.2 ml) and the enflurane concentration was determined with a sidestream rapid gas infrared analyzer (Datex Ultima anesthetic monitor; measuring error less than 0.2% in the range of 0.1–4% enflurane, calibrated before each investigation).

Data Analysis

Lung oxygen uptake, carbon dioxide excretion, and respiratory quotient were determined in 1-min intervals during steady-state conditions at an esophageal temperature of 28 degrees Celsius. In detail, the means of ten consecutive 1-min values during constant temperature and pump flow were determined and the results expressed as milliliters per minute. Lung V with dot O2was subsequently extrapolated to normothermic conditions assuming an increase of lung V with dot O2of 9% per degree centigrade. [14]Lung V with dot O2was then expressed as a percentage of whole-body V with dot O2.

With the institution of total CPB, oxygen uptake and carbon dioxide excretion decreased with decreasing temperature, reaching a steady state within 45 min at a constant esophageal temperature, as shown in Figure 1. In this period, oxygen uptake can be assumed to represent oxygen consumption. In this example, mean V with dot O2was 4.7, mean V with dot CO2was 3.5 ml/min, and the pulmonary respiratory quotient was 0.75 at an esophageal temperature of 28.2 degrees Celsius.

Similar recordings were obtained in all 16 patients. The means of 10 consecutive minute-to-minute determinations under conditions of constant esophageal temperature (28.4 degrees Celsius +/- 0.2) are presented in Table 2. Mean lung oxygen consumption was 5.3 +/- 1.6 ml/min, ranging from 2.8 to 9.9 ml/min. The results of the two patients with pulmonary hypertension (5.8 and 4.8 ml/min, respectively) were close to these values. Prebypass whole-body V with dot O2determined at nearly normothermic conditions (36.2 +/- 0.5 degrees Celsius) was 198 +/- 28 ml/min. Mean lung and mean whole-body respiratory quotient were similar at 0.84 +/- 0.09 and 0.77 +/- 0.09, respectively.

Extrapolating the measured lung V with dot O2of about 5–6 ml/min to 36 degrees Celsius by assuming an increase in pulmonary oxygen consumption of about 9% per degree centigrade suggests that the lung consumes about 11 ml oxygen per minute or about 5% of whole-body oxygen consumption for its own metabolic needs.

That the determined V with dot O2indeed represents lung V with dot O2is supported by the measured bronchial enflurane concentrations. Although the systemic blood was equilibrated in the membrane oxygenator with gas containing 2% enflurane, the end-tidal enflurane concentration was less than 0.1% in the bronchial system of all patients of the pilot study during total CPB. This suggests that the contribution of systemic (bronchial) blood flow to lung gas exchange during total CPB can be assumed to be negligible.

The human lung consumes about 5% of total-body oxygen uptake for its own metabolic requirements, as determined by indirect calorimetry during total CPB. This approach rests essentially on the assumption that pulmonary gas exchange is completely separated from body gas exchange during this period. Some possible limitations of this assumption must be considered because both bronchial circulation and transpulmonary gas exchange could have influenced our measurements.

In the normal lung, total bronchial blood flow is estimated to be about 1% of cardiac output, contributing a small amount to pulmonary capillary blood flow and gas exchange. [15]The major portion of this blood supplies the bronchial walls and the visceral pleura and is drained into the bronchial veins. We believe that the alveolar walls derive oxygen chiefly from the alveolar air, whereas the bronchi, the smaller air passages, and major portions of the visceral pleura use oxygen carried by the bronchial flow. [16,17]Bronchial blood flow in humans during total CPB has been assessed by measuring the amount of blood returning the left heart. Agostoni et al. found values of about 22 ml/min for healthy patients, [18,19]89 ml/min in patients with heart failure, [18]76 ml/min in patients with mitral stenosis, [19]and 40 ml/min in patients whose lungs are ventilated with dry air. [20]The same authors found a decrease in bronchial blood flow when alveolar pressure was increased. [21]Baile et al., [22]using the same approach, reported much higher values of about 140 ml/min (range, 8 to 1,043 ml/min). However, it remains unclear from these studies how much the bronchial circulation actually contributes to gas exchange during total CPB. From lung transplantation it is known that the bronchial circulation can be abolished without causing any obvious dysfunction in gas exchange, although there is much better healing of the tracheal-bronchial anastomosis when the bronchial arteries are anastomized. [23]Because our measurements based on respiratory gas analysis include only metabolic requirements met through direct diffusion of alveolar gas into the lung tissue, total lung oxygen consumption may have been underestimated because of the unknown proportion of need met by the bronchial circulation. However, because the detected bronchial enflurane concentration decreased to less than 0.1% with institution of total CPB, the contribution of the bronchial circulation to gas exchange during CPB should be negligible.

The second point to consider relates to transpleural diffusion. In isolated nonperfused dog lungs, measurable quantities of oxygen and carbon dioxide can traverse the visceral pleura. The absolute quantity varies with the magnitude of the concentration gradients. This can be derived from experiments in which the concentration gradient between extrapleural and inspired gas was increased from 6% to 12%, resulting in a parallel doubling of gas transfer. [9]In our investigation, 50% oxygen was used for both ventilating the lung and oxygenating the blood during total CPB, hence the transpleural partial pressure gradient for oxygen was minimized. In addition, the lungs in situ are completely covered by tissue, limiting gas exchange with ambient air. If a substantial amount of gas would have been lost or gained via the transpleural route, it is unlikely that oxygen and carbon dioxide are affected in the same way, so the respiratory quotient would not be within the physiologic range of 0.7 to 1. Because in our study mean lung and mean whole-body respiratory quotients were similar (0.84 and 0.77, respectively), the fraction of gas exchanged via the pleura during total CPB of the intact in situ lung should be negligible.

The third aspect relates to the effect of CPB and intraoperative conditions on the lung oxygen consumption of our study population, which consisted mainly of elderly men with marked cardiac disease. It is impossible from our data to specify whether and to which extent an open-chest, a fractional concentration of oxygen in inspired gas of 0.5, and a positive end-expiratory pressure of 3 mmHg might have influenced lung oxygen consumption. In addition, the absence of pulmonary artery blood flow during total CPB may have influenced the amount of oxygen consumed by the endothelium and pulmonary vascular smooth muscle. Thus our results may slightly underestimate lung oxygen consumption under physiologic conditions. Finally, measurements of lung oxygen consumption using respiratory gas analysis during total CPB may be influenced by the presence of atelectasis due to multifactorial causes. [24]We made efforts to minimize the development of atelectasis by ventilating the lungs of our patients continuously with 50% oxygen in air, including the administration of positive end-expiratory pressure. At no time were the lungs allowed to deflate.

Taking these considerations together, oxygen uptake and carbon dioxide excretion of the human lung during CPB are principally performed over the bronchial-alveolar system, probably with negligible contributions by transpleural diffusion and the bronchial circulation. Thus our determinations represent primarily lung VO2. However, we cannot exclude the possibility that they may have slightly underestimated lung oxygen consumption due to the development of atelectasis and the absence of pulmonary artery flow.

Much information on lung VO2has been received from lung slice preparations. There is a large variation in the measured values, ranging from 45 to 140 micro liter [center dot] min sup -1 [center dot] g sup -1 dry weight [25–27]in rat lung slices. However, this delicate technique delivers somewhat artificial results because rupturing of cell membranes and destroying tissue architecture can cause inactivation of normal biochemical activities and activation of others normally inhibited in intact cells. In addition, thickness of the slices [4]and location and method of slice resorting [5,27,28]markedly influence metabolic measurements in tissues (in the same studies, 6.1–9.1 micro liter [center dot] h sup -1 [center dot] mg sup -1 dry weight, and 1.1–1.4 micro liter [center dot] h sup -1 [center dot] mg sup -1 wet weight, rat lung slices). Therefore, results from these studies must be interpreted cautiously if extrapolated to the intact lung.

Further information on VO2derives from studies in isolated lungs or lobes of lungs with the advantage of intact cellular and tissue structures. Here values of 2–3 micro liter [center dot] h sup -1 [center dot] mg sup -1 dry weight [6,7]and 4.2 micro liter [center dot] h sup -1 [center dot] mg sup -1 wet weight, [8]or 0.76–0.98 ml/min in one left lung [9]were found in isolated dog lungs. However, it is unknown how the process of isolation and preparation and the ex situ perfusion technique may affect VO2under these conditions.

For in vivo assessments of lung VO2usually an indirect approach is used. Lung oxygen consumption has been estimated from the difference between VO2determined by indirect calorimetry and VO2determined by the Fick principle. While the latter is calculated as the product of systemic arteriovenous oxygen content difference and cardiac output and excludes by definition oxygen extraction of the lung, indirect calorimetry measures whole-body oxygen uptake. Because for this calculation different techniques with inherent imprecision are used (blood gas analysis, determination of hemoglobin, oxygen saturation, and cardiac output), the determined lung VO2has limited utility. Only if pulmonary oxygen consumption is greatly increased, such as in dogs with pneumococcal pneumonia (13–15% of whole-body VO2)[10]or in patients with far advanced pulmonary tuberculosis (12% of whole-body VO sub 2), [11]the differences may become large enough to exceed the measuring errors.

In a recent case report, lung VO2was determined by indirect calorimetry in a patient 2 days after double-lung transplantation requiring extracorporeal membrane oxygenation due to acute lung injury. [29]Although the measurements must be interpreted with caution because pulmonary artery blood flow probably was not completely inhibited, a similar VO2(7.8 ml/min) was reported.

In summary, we determined lung VO2during total CPB when lung gas exchange is separated from the systemic circulation. Extrapolating these data to 37 degrees Celsius suggests that the human lung consumes about 5% of total-body oxygen uptake.

Thierny DF: Lung metabolism and biochemistry. Ann Rev Physiol 1974; 36:209-31.
Euler V US, Liljestrand G: Observations on the pulmonary arterial blood pressure in the cat. Acta Physiol Scand 1946; 12:301-20.
Barron ESG, Miller ZB, Bartlett GR: Studies on biological oxidations. XXI. The metabolism of lung as determined by study of slices and ground tissue. J Biol Chem 1947; 171:791-800.
Levy SE, Harvey E: Effect of tissue slicing on rat lung metabolism. J Appl Physiol 1974; 37:239-40.
O'Neil JJ, Sanford RL, Wasserman S, Tierney DF: Metabolism in rat lung tissue slices: Technical factors. J Appl Physiol 1977; 43:902-6.
Faridy EE, Naimark A: Effect of distension on metabolism of excised dog lung. J Appl Physiol 1971; 31:31-7.
Weber KC, Visscher MB: Metabolism of the isolated canine lung. Am J Physiol 1969; 217:1044-52.
Bostroem B, Lochner W: Uber den Sauerstoffverbrauch der Lunge. Pflugers Arch 1955; 260:511-23.
Shepard JW Jr, Minh VD, Dolan GF: Gas exchange in nonperfused dog lungs. J Appl Physiol 1981; 51:1261-7.
Light RB: Intrapulmonary oxygen consumption in experimental pneumococcal pneumonia. J Appl Physiol 1988; 64:2490-5.
Fritts HW: Oxygen consumption of tissues in the human lung. Science 1961; 133:1070-2.
Phang PT, Rich T, Ronco J: A validation and comparison study of two metabolic monitors. J Parenter Enteral Nutr 1990; 14:259-61.
Takala J, Kleinanen O, Vaisanen P, Kari A: Measurement of gas exchange in intensive care: Laboratory and clinical validation of a new device. Crit Care Med 1989; 17:1041-7.
Harris EA, Seelye ER, Squire AW: Oxygen consumption during cardiopulmonary bypass with moderate hypothermia in man. Br J Anaesth 1971; 43:1113-20.
Deffebach ME, Charan NB, Lakshinarayan S, Butler J: The bronchial circulation: Small, but a vital attribute of the lung. Am Rev Respir Dis 1987; 135:463-81.
Bloomer WE, Harrison W, Lindskog GE, Liebow AA: Respiratory function and blood flow in the bronchial artery after ligation of the pulmonary artery. Am J Physiol 1949; 157:317-28.
Lilker ES, Nagy EJ: Gas exchange in the pulmonary collateral circulation of dogs. Am Rev Respir Dis 1975; 112:615-20.
Agostoni PG, Doria E, Bortone F, Antona C, Moruzzi P: Systemic to pulmonary bronchial blood flow in heart failure. Chest 1995; 107:1247-52.
Agostoni PG, Agrifoglio M, Arena V, Doria E, Sala A, Susini G: Systemic to pulmonary bronchial blood flow in mitral stenosis. Chest 1991; 99:642-5.
Agostoni PG, Arena V, Doria E, Susini G: Inspired gas relative humidity affects systemic to pulmonary bronchial blood flow in humans. Chest 1990; 97:1377-80.
Agostoni PG, Arena V, Biglioli P, Doria E, Sala A, Susini G: Increase of alveolar pressure reduces systemic to pulmonary bronchial blood flow in humans. Chest 1989; 96:1081-5.
Baile EM, Ling H, Heyworth JR, Hogg JC, Pare PD: Bronchopulmonary anastomotic and noncoronary collateral blood flow in humans during cardiopulmonary bypass. Chest 1985; 87:749-54.
Date H, Trulock EP, Arcidi JM, Sundaresan S, Cooper JD, Patterson A: Improved airway healing after lung transplantation. J Thorac Cardiovasc Surg 1995; 110:1424-33.
Hachenberg T, Tenling A, Hansson HE, Tyden H, Hedenstierna G: Ventilation-perfusion inequality in patients undergoing cardiac surgery. Anesthesiology 1994; 80:509-19.
Caldwell PRB, Wittenberg BA: The oxygen dependency of mammalian tissue. Am J Med 1974; 57:447-52.
Krebs HA: Body size and tissue respiration. Biochim Biophys Acta 1950; 4:249-69.
Engelbrecht FM, Maritz G: Influence of substrate composition on in vitro oxygen consumption of lung slices. S Afr Med J 1974; 48:1882-4.
Massaro GD, Gail DB, Massaro D: Lung oxygen consumption in mitochondria of alveolar epithelial and endothelial cells. J Appl Physiol 1975; 38:588-92.
Lee KH, Rico P, Boujoukos AJ, Keenan RJ, Pinsky MR: Measurement of lung oxygen consumption in a patient after double-lung transplant. J Thorac Cardiovasc Surg 1995; 110:1764-5.