Chest Computed Tomography in the Acute Respiratory Distress Syndrome
Special Feature
Chest Computed Tomography in the Acute Respiratory Distress Syndrome
By David J. Pierson, MD, FACP, FCCP
The acute respiratory distress syndrome (ards) is one of the most important conditions managed by intensivists. Both our understanding of the genesis and natural history of this syndrome and our approach to its management have changed dramatically in the last several years. This brief essay outlines the ways in which knowledge about ARDS and its treatment have been advanced by studies using chest computed tomography (CT)1,2 and offers recommendations for the use of CT in the assessment and management of individual patients.
Is ARDS a Homogeneous Process?
ARDS was first described more than 30 years ago as a form of diffuse lung injury that followed any of a variety of insults. Pathologically, it was characterized by widespread atelectasis, alveolar hemorrhage, and edema, and eventually also by the formation of hyaline membranes and fibrosis. A requirement for diagnosis of the syndrome was the presence of diffuse bilateral alveolar opacities on the standard chest x-ray. Portable chest radiographs in most patients with ARDS showed a pattern consistent with pulmonary edema, usually with a normal-sized heart, especially early in the syndrome’s clinical course. Pathologists, radiologists, and clinicians alike thought of ARDS as a diffuse process affecting the lungs more or less homogeneously.
However, in 1986, investigators in Milan and also in Seattle demonstrated that, when examined by chest CT, ARDS was anything but homogeneous. In 10 patients, Gattinoni et al3 demonstrated a gradient of CT density from the uppermost (ventral) to the lowest (dorsal) lung regions when patients were in the supine position. When positive end-expiratory pressure (PEEP) was applied incrementally in these patients, CT scanning showed that overall lung volume increased and lung density progressively decreased. However, improved oxygenation with increasing PEEP could not be predicted from these morphological changes in individual patients. In the same year, Maunder et al showed in 13 patients with ARDS that, although the standard chest x-ray tended to show a more or less homogeneous pattern of lung opacity, the pattern on CT scan revealed marked regional heterogeneity.4 Some lung areas appeared normal, while others were characterized by patchy infiltrates.
Subsequent studies have shown that no areas of the lungs in patients with ARDS are completely normal. What varies in different areas is lung density, influenced prominently by gravitational forces in the presence of an increase in lung fluid.1 Compression atelectasis, a loss of gas in the most dependent regions of the lung, occurs in normal individuals under anesthesia and paralysis. However, whereas compression atelectasis comprises perhaps 10% of the lung in anesthetized normal persons, it can occupy as much as 70-80% of the lung fields in severe ARDS. The amount of aerated lung participating in gas exchange may thus be reduced to 20-30% of the normal amount. This observation prompted Gattinoni to coin the term "baby lung,"5 which emphasizes the point that the lungs in patients with ARDS are not so much stiff as they are small in comparison to the normal condition. Lung compliance is reduced not so much because of lung stiffness as because of a reduction in effective lung size.
The observation from chest CT scans that effective, ventilated lung volume in ARDS is markedly reduced has helped to link the findings of animal studies on ventilator-induced lung injury to what is seen in patients. Studies in a variety of animal models have demonstrated that mechanical ventilation with tidal volumes in the range of 40-60 mL/kg produces an acute severe lung injury that is similar in some ways to that seen in ARDS. The use of tidal volumes of 10-15 mL/kg in patients with ARDS whose available, ventilated lung volume has been reduced to only 20-30% of normal may subject those areas of their lungs to distending volumes similar to what has been used in the animal models.
The pattern of lung density revealed by CT scanning in patients with ARDS (particularly in the early phase of the syndrome) typically reveals three distinct zones distributed vertically in the thorax in the supine position.1,2 Closest to the sternum (that is, uppermost) is a lucent zone that appears grossly normal in density. Closest to the spine (that is, most dependent) is a dense zone of apparent collapse and consolidation. Between these is a zone of lung of intermediate density, believed to represent collapsed alveoli that potentially could be inflated (that is, recruited). Lung weight in ARDS is increased two- or threefold, and this weight is accounted for by the increased tissue volume and corresponding decrease in gas volume.
Puybasset et al studied 21 patients with acute lung injury in comparison with 10 normal volunteers.6 Their findings support Gattinoni et al’s hypothesis that ARDS causes small rather than stiff lungs, and found that the volume reduction was mainly in the lower lobes. Total lung volume (aerated vs nonaerated lung) was reduced 27% in their patients, with the upper lobes having 99% and the lower lobes 48% of the normal values.
Although the uppermost lung zones in patients with ARDS have been described as normal in appearance, CT studies by Gattinoni et al have shown that even these areas are in fact abnormal.1 A recent study showed that ground-glass opacification was present in all areas of the lungs in patients with ARDS, while consolidation was predominantly dorsal and caudal in the thorax.7 Gattinoni et al explain the apparent paradox of diffuse, homogeneous lung involvement in the face of a well-established ventral-to-dorsal CT density gradient as follows.1,2 The homogeneously affected, heavy lung exists in a gravitational field, with hydrostatic pressures transmitted through the lung as if it were liquid. Gas is squeezed out of the lung in its lowermost areas much like air is squeezed out of a saturated sponge.
"Pulmonary" vs. "Extrapulmonary" ARDS
ARDS is a pulmonary disorder, but whether it stems from a primary lung process or from some extrapulmonary disease in an individual patient may make an important difference. Here again our thinking has been influenced prominently by investigations by Gattinoni et al in Milan. In a physiologic study of 21 patients, they found a number of differences between the 12 individuals with ARDS originating from pulmonary disease (ARDSp) and the nine with ARDS in the setting of extrapulmonary disease (ARDSexp).8 Although total respiratory system elastance and end-expiratory lung volume off PEEP were the same in the two groups, patients with ARDSp had higher lung elastance and relatively normal chest wall elastance, whereas the reverse was the case in patients with ARDSexp. The increased chest wall elastance in the patients with extrapulmonary ARDS correlated highly with increased intra-abdominal pressures in these patients. Further distinguishing between ARDSp and ARDSexp in a CT-based study, Goodman et al found that the former tended to be asymmetric with a mix of ground-glass opacification and consolidation, whereas ARDSexp was characterized mainly by ground-glass opacification (i.e., edema and collapse) throughout the lungs.7
Documenting the Effects of PEEP
If the lung in ARDS behaves like a saturated sponge, with collapse of the most dependent areas as a result of increased hydrostatic pressure, then it makes sense to apply PEEP equal in cm H2O to the hydrostatic pressure column corresponding to the vertical height of the lung. Early studies by Gattinoni et al showed that the CT density of individual lung areas increased with increasing PEEP of approximately this range of pressures.9,10 Unfortunately, more ventral areas of lung are subjected to the same PEEP as those targeted in the dependent areas, and the former are at risk of overdistension.
It would therefore be of value to be able to identify alveolar overdistension and to distinguish between successful alveolar recruitment and this undesirable effect of the incremental application of PEEP. Vieira et al11 studied six patients with acute lung injury in comparison with six healthy individuals. They determined that a CT number of -900 Houndsfield units (HU) or less corresponded to lung overdistension, and proposed that this threshold be used to distinguish between PEEP-induced alveolar recruitment and lung overdistension.
Vieira et al subsequently studied lung CT patterns in patients with acute lung injury as a function of whether they demonstrated a lower inflection point (LIP) on the pressure-volume curve of the total respiratory system.12 They studied eight patients with and six patients without a demonstrable LIP. Total respiratory system and lung compliances were lower in patients who demonstrated a LIP, whereas no other respiratory parameters measured differed in the two groups. The application of PEEP induced lung recruitment associated with alveolar overdistension, as demonstrated by lung density histograms, only in patients who did not demonstrate a LIP. According to the findings of this study, air and tissue are more homogeneously distributed in the lungs of patients demonstrating a LIP, and in such patients increasing PEEP recruits additional alveoli without producing lung overdistension.
Effects of Body Position Changes
The increased hydrostatic pressure in the most dependent parts of the lung, due to the increased edema there, produces alveolar compression and allows the dependent lung to develop atelectasis.1 PEEP acts as a counterforce to keep open the previously collapsed alveoli and thus reduces the amount of unventilated lung tissue. Turning the patient into the prone position apparently accomplishes the same thing.
In several papers Gattinoni et al reported the effects of moving patients with ARDS from the supine to the prone position, as shown by chest CT.13,14 These studies showed that, within just a few minutes of moving the patients into the prone position, the most dense lung areas moved from dorsal to ventral regions. The edema itself did not shift position; protein-rich interstitial edema is not free to move. Instead, the observed density redistribution results from squeezing air out from the sponge-like lung in the dependent regions because of superimposed gravitational pressure.2
It has been observed for years that turning ARDS patients into the prone position improves arterial oxygenation, at least in many cases. Gravitational forces and the "sponge lung" effect partly explain this phenomenon, but are not the complete story. In addition, a higher PO2 does not necessarily a better outcome make, as the disappointing clinical trials of nitric oxide in ARDS have shown. A large-scale multicenter Italian trial of prone positioning in acute respiratory failure is nearing completion, and may provide the much-needed answer to the question of whether this maneuver affects mortality, the duration of mechanical ventilation, or some other outcome.
Clarifying the Natural History of ARDS
Most of the information summarized above about lung density patterns in ARDS was derived from patients in the first few days of the syndrome. Early on, ARDS is typically manifested by hypoxemia—that is, by right-to-left shunt and low ventilation-perfusion (V/Q) areas—and CO2 elimination is not usually a major problem. Although patients with severe ARDS may still have oxygenation failure after two or more weeks, the typical functional pattern is one of ventilation, not oxygenation, problems—that is, high V/Q regions and increased dead space. Late stage tends to be characterized by low static compliance, hypercapnia, less difficulty in maintaining arterial oxygenation, and a diminished requirement for PEEP.15
These differences between early- and late-stage ARDS have structural counterparts, as revealed by chest CT. While ground-glass opacification characteristic of pulmonary edema and areas of consolidation are usual in the early stage, findings reminiscent of emphysema, including bullae, typically appear as the syndrome persists. In one study, pneumothorax occurred more often in late ARDS (that is, after > 2 weeks of mechanical ventilation) than in either intermediate (1-2 weeks) or early (< 1 week) ARDS.15
Chest CT has also provided insight into the structural sequelae of ARDS in survivors. In a study of 27 ARDS survivors who underwent chest CT, the most common abnormality (in 85% of patients) was a reticular pattern, which was significantly more prevalent anteriorly in the lung than posteriorly.16 In this study, the extent of this reticular pattern on follow-up CT was significantly correlated with the duration of mechanical ventilation, most prominently so in patients who had undergone pressure-controlled inverse-ratio ventilation.
Role of Chest CT in Managing Patients with ARDS
Several helpful management principles have come out of studies using CT. Enough PEEP should be used in patients with ARDS to open the lung. While constructing a pressure-volume curve in the individual patient is impractical and usually requires paralyzing the patient, studies using these curves in combination with CT imaging indicate that at least 10 cm H2O of PEEP is needed in most instances. Higher levels will likely be necessary in patients with the most severe oxygenation deficits. Lung overdistension should be avoided, however, and this is best accomplished by limiting delivered tidal volume to less than 6 mL/kg. Finally, management in the late phase of the illness may need to be different than in its initial days.
The above principles apply to the management of ARDS in general. Does chest CT have a role in managing the individual patient? With respect to most of the concepts discussed in this essay, probably not. That is, CT is not needed to make the diagnosis (in fact, the latter relies on standard portable chest radiographs, and no accepted corresponding set of CT findings exists). It would be impractical and expensive to perform chest CT scans routinely to assess the effects of PEEP, although there could come a time when modified, less expensive CT scanners would be available for bedside use in the ICU. As long as the CT scanner resides some distance from the ICU, obtaining images requires transporting an often critically ill patient, and this poses real hazards as well as inconvenience for staff.
However, the risks and hassles of patient transport are probably worth it in some clinical circumstances. Bedside chest x-rays are insensitive and often poorly reproducible. When patients develop new fever and other signs of sepsis in the absence of a new parenchymal change on chest radiograph, chest CT can be helpful in identifying unsuspected fluid collections, abscesses, or areas of lung necrosis. Chest CT may be particularly helpful in evaluating patients with persistent ARDS following surgery or trauma involving the chest. Assessing the completeness of evacuation of pleural empyema usually requires CT scanning, and this is particularly important if the patient is not responding well to treatment. Patients late in the course of ARDS who have repeated episodes of pneumothorax may be physiologically compromised by loculated pleural air collections that are not apparent on the bedside chest radiograph. CT-guided catheter drainage of such air collections may be helpful,17 as is also the case with difficult-to-reach pleural or intraparenchymal fluid collections that may be infected.
Summary
Present knowledge of the pathophysiology and natural history of ARDS has been advanced substantially by studies using chest CT. Although the field remains in evolution, cogent explanations based in part on CT imaging now characterize the structural patterns of different forms of ARDS and also differences between early and late stages of the syndrome. CT-derived data, in combination with the findings of studies using animal models, have led to a new approach to ventilator management in ARDS, which is gaining increasing support from clinical trials. The CT scan can provide the clinician with valuable diagnostic information in assessing patients with ARDS who are not improving or who are experiencing barotrauma or infectious complications.
References
1. Pelosi P, et al. Computed tomography in adult respiratory distress syndrome: What has it taught us? Eur Respir J 1996;9(5):1055-1062.
2. Gattinoni L, et al. The acute respiratory distress syndrome. In: Albert RK, et al, eds. Comprehensive Respiratory Medicine. London/Philadelphia: Mosby; 1999:69.1-69.16.
3. Gattinoni L, et al. Morphological response to positive end expiratory pressure in acute respiratory failure. Computerized tomography study. Intensive Care Med 1986;12(3):137-142.
4. Maunder RJ, et al. Preservation of normal lung regions in the adult respiratory distress syndrome. Analysis by computed tomography. JAMA 1986; 255:2463-2465.
5. Gattinoni L, Pesenti S. ARDS: The dishomogeneous lung: Facts and hypothesis. Intensive Crit Care Digest 1987;6:1-4.
6. Puybasset L, et al. A computed tomography scan assessment of regional lung volume in acute lung injury. The CT Scan ARDS Study Group. Am J Respir Crit Care Med 1998;158(5):1644-1655.
7. Goodman LR, et al. Adult respiratory distress syndrome due to pulmonary and extrapulmonary causes: CT, clinical, and functional correlations. Radiology 1999;213(2):545-552.
8. Gattinoni L, et al. Acute respiratory distress syndrome caused by pulmonary and extrapulmonary disease. Different syndromes? Am J Respir Crit Care Med 1998;158(1):3-11.
9. Gattinoni L, et al. Regional effects and mechanism of positive end-expiratory pressure in early adult respiratory distress syndrome. JAMA 1993;269(16): 2122-2127.
10. Pelosi P, et al. Vertical gradient of regional lung inflation in adult respiratory distress syndrome. Am J Respir Crit Care Med 1994;149(1):8-13.
11. Vieira SR, et al. A lung computed tomographic assessment of positive end-expiratory pressure-induced lung overdistension. Am J Respir Crit Care Med 1998; 158(5):1571-1577.
12. Vieira SR, et al. A scanographic assessment of pulmonary morphology in acute lung injury. Significance of the lower inflection point detected on the lung pressure-volume curve. Am J Respir Crit Care Med 1999;159(5):1612-1623.
13. Langer M. The prone position in ARDS patients. A clinical study. Chest 1988;94(1):103-107.
14. Gattinoni L, et al. Body position changes redistribute lung computed-tomographic density in patients with acute respiratory failure. Anesthesiology 1991; 74(1):15-23.
15. Gattinoni L, et al. Lung structure and function in different stages of severe adult respiratory distress syndrome. JAMA 1994;271(22):1772-1779.
16. Desai SR, et al. Acute respiratory distress syndrome: CT abnormalities at long-term follow-up. Radiology 1999;210(1):29-35.
17. Chon KS, et al. CT-guided catheter drainage of loculated thoracic air collections in mechanically ventilated patients with acute respiratory distress syndrome. AJR Am J Roentgenol 1999;173(5):1345-1350.
Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.