IN this issue of Anesthesiology, Thille and an international group of coworkers recognized for their expertise in acute respiratory distress syndrome (ARDS) bring convincing evidence that positive end-expiratory pressure-induced alveolar recruitment does not differ between patients with pulmonary and extrapulmonary ARDS.1The authors recommend that the origin of ARDS should not influence selecting the appropriate positive end-expiratory pressure level: The recruitability of the lungs seems similar in pneumonia, aspiration, inhalation injury, alveolar hemorrhage, pulmonary contusion and secondary lung injury resulting from sepsis, acute pancreatitis, multiple trauma, cardiopulmonary bypass, and massive transfusion.
This multicenter study contradicts the classic and widespread belief that ARDS from primary pulmonary causes is less responsive to positive end-expiratory pressure, prone position, and recruitment maneuvers than ARDS from extrapulmonary causes2and outlines some of the methodologic bias originally involved in generation of this belief. Initially, two studies suggested that lung recruitment obtained by increasing intrathoracic pressure resulting from prone positioning was substantially lower in patients with direct injury to the lung compared with patients with secondary lung injury.2,3In both studies, recruitment was measured using the quasi-static compliance method, a method that entails a risk of underestimating alveolar recruitment and changes in respiratory mechanics, as pointed out by Thille et al. 1In the second study reporting different arterial oxygen response to prone positioning between primary and secondary ARDS, the difference in arterial oxygenation, although statistically significant, was of small magnitude and of questionable clinical significance.3In addition, questions can be raised regarding the statistical analysis.3A third study, performed in five patients with ARDS caused by severe pneumonia and five patients with extrapulmonary ARDS, reported that lung recruitment after three consecutive sighs was considerably less in the former than in the latter.4The method for measuring lung recruitment, however, could be questioned: End-expiratory lung volume was measured by the closed-circuit helium dilution method before and after sighs, and lung recruitment was computed as the sigh-induced change in end-expiratory lung volume, ignoring whether the increase in lung volume was related to (over)inflation of previously aerated lung or to lung recruitment. Another clinical study demonstrated different gas exchange response to nebulization of prostacyclin between primary and secondary ARDS.5In six patients with severe pneumonia resulting from infection or aspiration, nebulized prostacyclin decreased arterial oxygenation without changing pulmonary arterial pressure. In contrast, in nine patients with lung injury resulting from extrapulmonary sepsis, necrotizing pancreatitis, or multiple trauma, it significantly improved arterial oxygenation and decreased mean pulmonary arterial pressure. Differences in gas exchange response, however, coincided with differences in lung morphology assessed by lung computed tomography. Primary ARDS had a diffuse loss of lung aeration, whereas secondary ARDS had a focal lung aeration predominating in the lower lobes and sparing the upper lobes. Differences in lung morphology rather than the cause of lung injury likely explain the difference found in gas exchange response because it has been clearly demonstrated that diffuse loss of lung aeration is not specific to pneumonia: Mild forms remain focally distributed, whereas severe forms affect lung tissue diffusely.6,7
After the initial report putting forward the idea that pulmonary and extrapulmonary ARDS behave differently in terms of alveolar recruitment and should require specific clinical management,2many contradictory studies were published.6,8–14Although a higher incidence of lung consolidation and focal loss of aeration is observed in secondary ARDS,6,15differences in lung morphology (focal vs. diffuse loss of lung aeration) do not exactly coincide with the cause of lung injury.6,10Several studies have reported that alveolar recruitment resulting from increases in positive end-expiratory pressure,9,12,13recruitment maneuvers,8and prone positioning11were not influenced by the cause of lung injury at the early phase of ARDS. Interestingly, in all of these studies, lung recruitment was measured from static pressure-volume curves, and in their study, Thille et al. 1elegantly demonstrate that the quasi-static compliance method is inappropriate for measuring recruitment. Finally, Gattinoni et al. 14recently reported a series of 68 ARDS patients in whom computed tomography of the whole lung was performed at airway pressures of 5, 15, and 45 cm H2O. Contradicting their initial hypothesis,2patients with primary ARDS had a higher percentage of recruitable lung than patients with secondary ARDS.
Another confusing factor is the difficulty of classifying ARDS in one or the other category. In the study by Thille et al. ,1physicians well known for their expertise in ARDS were unable to classify 37% of the patients. The reasons are multiple. Lung infection rapidly complicates the course of mechanical ventilation, inducing a direct injury to the lung. In anesthetized and ventilated animals, disseminated foci of bronchopneumonia are found only a few hours after the initiation of mechanical ventilation.16,17In the study of Thille et al. , 45% of the patients were included at a late phase of ARDS, a condition that complicates classification. Over time with mechanical ventilation, the risk of lung superinfection rapidly increases and extensive fibrosis may occur, contributing to the mixing of pulmonary and extrapulmonary lung injuries. Conversely, a focal bronchopneumonia, initially limited to the lung parenchyma, may induce secondary septic shock and extrapulmonary lung injury. In patients with multiple trauma, it is not easy to discriminate pulmonary contusion from lung injury resulting from hemorrhagic shock and massive transfusion. Finally, as previously reported,7it can be difficult to discriminate pulmonary and extrapulmonary ARDS because both of them may coexist in the same patient.
The “primary/secondary ARDS story” illustrates the critical importance of using adequate methodology in physiologic human studies. Inaccurate methodology may lead to incorrect interpretation of the data and false theories. When, in addition, the theory flatters and coincides with good sense—it appears quite logical that a consolidated infected lung recruits less than a lung with alveolar-interstitial edema—then, despite numerous studies unable to confirm the theory, the belief persists for a long time. One should always remember what Mahatma K. Gandhi said in 1921: “An error does not become truth by reason of multiplied propagation, nor does truth become error because nobody will see it.”18
Anesthesiology and Critical Care Medicine, University of Paris 6, Paris, France. jjrouby.pitie@invivo.edu