Thank you for your interest in our article.1We read with great interest Dr. Russell's study reporting beneficial effects of intermittent positive airway pressure on the nonventilated lung in patients suffering from hypoxemia (i.e. , pulse oximetry less than 95%) during one-lung ventilation (OLV).2Oxygen administered into the nondependent lung should be used to treat and prevent hypoxemia during OLV. As mentioned in our article,1oxygen may be used in different ways on the nondependent lung with or without continuous positive airway pressure and transient reinflation of the collapsed lung without interfering with surgery. The method described by Russell et al. seems very interesting.2These authors reported that slow and very brief inflation with pure oxygen is sufficient to treat hypoxemia. However, the efficiency of this simple technique depends on the importance of the shunt-like effect and hypoxemia that requires treatment.3,4
We completely agree with Drs. Tripathi and Papadimos that oxygen saturation measured by pulse oximetry (Spo2) below 90% is an arbitrary limit. We are also aware that hypoxemia does not necessarily involve hypoxia and organ injury, as you so pertinently point out. However, this is an educational article and we feel the need to have a clear message. In order to refer to hypoxemia a threshold must be established; 90% seems to be a commonly accepted, logical, and adequate limit.*Obviously, as with all constants and medical situations, the clinical context holds the same importance as the definition itself.
Crossing the limit is as noteworthy as the manner in which it is decreasing, sometimes sharply and rapidly. In addition, although the patient's tolerance may remain satisfying, any new incident has the potential for dramatic consequences given a lack of margin. Consequently, it appears reasonable for us to take action as soon as Spo2drops below 90%.
We completely agree that a lower Spo2may be acceptable in patients suffering from acute respiratory distress syndrome. We have had similar experiences in our center among patients with acute respiratory distress syndrome who have refractory hypoxemia and are undergoing veno-venous extracorporeal membrane oxygenation. The arterial oxygen saturation is only one component for oxygen delivery.5,6However, it seems difficult to test, within a randomized controlled trial, the hypothesis that Spo2below 80% would be safe during anesthesia in planned surgery. Most surgical patients are prepared for surgery. Therefore, hemoglobin concentration and cardiac status are known and deemed satisfying before the intervention. Nonetheless, both may require monitoring and correction to optimize oxygen delivery to tissues.7,8
As mentioned in your response to our manuscript, “vigilant consideration must be given to a myriad of parameters that need to be perfectly controlled before discontinuation of surgical procedure.” Certain selected patients with hypoxemia may benefit from extracorporeal gas exchange in order to undergo thoracic surgery.9
However, it seems important not to avoid overcomplication; pulmonary catheter, mixed venous oxygen saturation, and continuous cardiac output do not qualify as day-to-day monitoring in ordinary thoracic surgery. Based on our belief and experience, an OLV situation where the Spo2is below 90% with a fraction of inspired oxygen of 100% first must always be tagged as an unusual and potentially dangerous situation, requiring, if not an immediate pause in the procedure, at least more acute attention from the anesthesiologist and understanding of the hidden physiopathologic mechanism behind this hypoxemia, with the specificities of OLV.
We agree with Gallart et al. that the chapter's title “Pharmacological Treatment to Limit Hypoxic Pulmonary Vasoconstriction” could lead to confusion. Consequently, the following title “Pharmacological Treatment to Limit the Shunt-like Effect” might be better. Regarding the effects of inhaled nitric oxide (iNO) on oxygenation during OLV, the improvement of oxygenation is not systematic. Indeed, iNO improves oxygenation by reducing intrapulmonary shunt in patients with severe hypoxemia during OLV.10,11This hypoxemia is correlated with the lung perfusion ratio between the operated lung and the ventilated lung; Pao2decreased proportionally with perfusion of the operated lung.4In this situation, iNO improves oxygenation by increasing the perfusion of the ventilated lung without modifying hypoxic pulmonary vasoconstriction. Concerning positive end-expiratory pressure, it may paradoxically impair oxygenation by shifting pulmonary blood flow to the nondependent lung, thereby increasing intrapulmonary shunt.12Indeed, increasing airway pressure in the dependent lung is responsible for high pulmonary vascular resistances because of the compression of the intraalveolar vessels. Thus positive end-expiratory pressure in the dependent lung may divert blood flow away from the ventilated lung. This may counteract hypoxic pulmonary vasoconstriction in the nondependent lung and result in an increase in the pulmonary shunt fraction. In this specific situation, iNO may be useful in its ability to reduce the shunt while a positive end-expiratory pressure is applied to ventilated lung during OLV.13We agree that almitrine during OLV must be administered at a low dose and during a short period of time, but potential toxicity needs to be known. In conclusion, almitrine as well as iNO should not be routinely used, however, but should be considered when other strategies fail to improve oxygenation.
†Haut Leveque, Bordeaux University Hospital, Bordeaux, France. firstname.lastname@example.org