HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) All Versions of this Article: 2232010721v1 223/2/468    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Padley, S. P. G. Articles by Hansell, D. M. Search for Related Content PubMed PubMed Citation Articles by Padley, S. P. G. Articles by Hansell, D. M.

Asymmetric ARDS Following Pulmonary Resection: CT Findings—Initial Observations¹

¹ From the Departments of Radiology (S.P.G.P., D.M.H.), Surgery (P.G.), Critical Care Medicine (S.J.J.), and Respiratory Medicine (A.U.W.), Royal Brompton Hospital, Fulham Rd, London SW3 6NP, England. Received April 4, 2001; revision requested May 23; revision received August 13; accepted September 28. Address correspondence to S.P.G.P. (e-mail: s.padley@ic.ac.uk).

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
PURPOSE: To investigate whether asymmetric distribution of acute respiratory distress syndrome (ARDS) following lobectomy is due to compensatory hyperexpansion of the residual lung within the hemithorax operated on and to discern if this distribution reflects true asymmetry of the disease process.

MATERIALS AND METHODS: Retrospective review of the intensive care unit database was performed over a period of 6 years to identify all cases of lung injury following lung surgery that satisfied the American-European consensus criteria for ARDS. Time to onset following surgery, time of subsequent computed tomographic (CT) examination, patient age and sex, and nature of surgery were recorded, as well as eventual patient status (ie, death or discharge). Availability of both preoperative and postoperative CT scans was required for inclusion for further analysis. These images were analyzed on a commercial CT workstation for the volume of lung resected and the pre- and postoperative volume and density of each lung. Expected postoperative densities (preoperative density adjusted for volume) were compared with observed postoperative densities.

RESULTS: Review disclosed 583 patients who underwent lobectomy or segmentectomy. Seventeen patients (2.9%) developed postoperative ARDS. Nine of these patients had pre- and postoperative CT scans available for analysis. In eight of nine cases, density increased more in the nonoperated lung than in the operated lung (P = .01). The degree of density increase in the nonoperated lung was significantly greater (305 mg/mL; range, 48–449 mg/mL) than that in the operated lung (13 mg/mL; range, -198 to 231 mg/mL; P < .001).

CONCLUSION: Following lobectomy, there appears to be a truly asymmetric form of ARDS rather than compensatory hyperexpansion of the residual lung on the operated side.

© RSNA, 2002

Index terms: Lung, injuries, 60.413, 60.452, 60.453 • Lung, surgery, 60.413, 60.452, 60.453 • Respiratory distress syndrome, adult (ARDS), 60.413

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Acute respiratory distress syndrome (ARDS) is a recognized complication of lung resection and has been previously termed postpneumonectomy pulmonary edema (PPE) (1). When ARDS is described following lobectomy, it may similarly be termed postlobectomy edema (PLE). ARDS complicates approximately 5% of lung or lobar resections—the reported frequency ranges from 4% to 15% (2–5). Although the exact definition of PPE varies between series (1–3,6), the defining features of ARDS are common to all descriptions, such that in its most severe form, PPE is indistinguishable from ARDS with regard to both clinical manifestation and pathologic features (2).

In 1992, an American-European Consensus Committee on ARDS produced a detailed definition of diagnostic criteria. Central to this definition was the development of inflammation and increased capillary permeability, associated with a wide number of physiologic, clinical, and radiologic abnormalities that cannot be explained by elevated pulmonary venous pressure (7). This combination of features was termed acute lung injury, with ARDS being the preferred term to describe more severe disease. The diagnosis is usually made after the patient has developed refractory arterial hypoxemia, which is accompanied by the development of typically bilateral pulmonary opacities on chest radiographs.

We have observed that chest radiographs and computed tomographic (CT) scans of patients admitted to the intensive care unit following the development of postlobectomy acute lung injury or ARDS may manifest an asymmetric pattern of lung involvement. Unlike almost all other causes of ARDS, which typically result in symmetric bilateral changes (with the exception of the expected anteroposterior attenuation gradient), patients with postlobectomy ARDS appear to have relative sparing of the lung that underwent lobectomy. Thus, the purpose of our study was to investigate whether an asymmetric distribution of ARDS following lobectomy is due to compensatory hyperexpansion of the residual lung within the hemithorax operated on and to discern if this distribution reflects true asymmetry of ARDS.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
The intensive care database was reviewed by one author (S.J.J.) to identify all cases of lung injury following lung surgery that satisfied the consensus definition of ARDS (ie, acute onset of hypoxia, with a partial pressure of oxygen [arterial] to fraction of inspired oxygen ratio of less than 200 mm Hg, a pulmonary artery occlusion pressure of less than 18 mm Hg, and radiologic evidence of pulmonary opacities on chest radiographs) from the time of database inception in May 1993 to June 1999. At our institution, ethics committee rules do not require informed consent from patients for retrospective review of records for research purposes, as long as anonymity is guaranteed.

Patients included in the database who developed ARDS following lobectomy or segmentectomy were identified. The timing of ARDS onset following surgery was recorded. Additional data recorded included patient age and sex, site of operation, and status at time of discharge. Only those patients were included in the series in whom both preoperative CT scans (obtained within 1 month of surgery) and postoperative CT scans (obtained during the period of ARDS) were available. These images were retrieved from the electronic archive and reloaded onto a commercial workstation (Magicview; Siemens, Erlangen, Germany). All CT examinations were performed on an electron-beam CT scanner (Imatron, San Francisco, Calif). Preoperative scans were obtained as part of a staging examination for presumed bronchogenic neoplasia (volume acquisition followed intravenous contrast material administration, with 6-mm collimation and a 6-mm reconstruction interval). Postoperative scans were obtained with a thin-section rapid-acquisition technique (3-mm collimation at 10-mm intervals; high spatial frequency reconstruction algorithm; acquisition time, 100 msec) as part of a standardized protocol for management of ARDS.

Image Analysis
All image analysis was undertaken by a single author (S.P.G.P.).

Volume measurements.—With standard, commercial volume-rendering software (Magicview 1000 VA 31, release C; Siemens), the preoperative volumes of the left and right lungs and the volume of the lobe to be resected were determined. The lung–chest wall interface was delineated with the track ball at each level, and the total volume of consecutive levels was automatically calculated to provide a lung volume measurement. A similar procedure was repeated for the lobe to be resected, as determined by segmental airway and fissural anatomy. In the two patients who underwent left upper division resection with preservation of the lingular segments, a "best estimate" of the boundaries of the resected lung was derived from visualization of the segmental branching pattern to estimate the region of the lung supplied by the lingular segments. The process of volume determination for each lung and each resected lobe or segment was repeated three times, and the mean value was used for subsequent calculation. This repetition was performed to negate small errors in volume tracing. The postoperative thin-section CT scans were analyzed in a similar manner to determine the total postoperative volumes of the lung that underwent surgery and the lung that did not undergo surgery (hereafter referred to as the "operated lung" and the "nonoperated lung," respectively). The postoperative volume determination made allowance for the noncontiguous nature of the images, since the z-axis dimension was derived from table position data.

Preoperative density measurement.—Tissue density measurements in milligrams per milliliter (equivalent to the HU measurement + 1,000) (8) were determined in the anterior and posterior portions of both lungs at preoperative examination. Density measurements were assumed to bear a linear relationship to lung volume for a given lung, as demonstrated by results of previous studies (9–11). Density measurements were determined by placement of 957-pixel regions of interest (ROIs) in the operated lung. These ROIs consisted of an anteroposterior pair within the lung subsequently retained following lobectomy. A further pair of ROIs were placed in a similar manner within the nonoperated lung; these mirrored the pair placed in the operated lung, both for craniocaudal level and anteroposterior position. From each anteroposterior pair, an average attenuation value was obtained, and this was used for subsequent analysis. Care was taken at the time of ROI placement to avoid large or medium-sized vessels, although inclusion of some pulmonary vessel branches was inevitable.

Postoperative density measurement.—Similarly, postoperative density measurements were obtained by placing an anteroposterior pair of ROIs in an anatomic location (as judged by airway and fissural anatomy) comparable to those measured prior to operation in the portion of the lung not resected, with a second pair of ROIs placed at the same level in the nonoperated lung (Fig 1). Wherever necessary, minor positional differences in ROI placement were made to allow for postoperative anatomic adjustment. The placement of ROIs was also adjusted to avoid areas of parenchymal consolidation or pleural fluid collections. At the time of CT analysis, a visual assessment of the proportion of each lung occupied by intense opacification was made to the nearest 5%. This subjective visual method has been previously validated and used for the formal quantitative assessment of fibrosing lung disease (12).

View larger version (142K): [in this window] [in a new window] [Download PPT slide]   Figure 1a. Transverse CT images obtained before and after surgery. (a) Two ROIs are shown in the posterior lung parenchyma. This patient subsequently underwent left upper lobectomy. The anterior pair of ROIs are not demonstrated. (b) Image shows postoperative development of ARDS, with two posterior ROIs positioned at an equivalent anatomic location. Again, the anterior pair of ROIs are not demonstrated, since the workstation is able to display only two ROIs at a time.

View larger version (151K): [in this window] [in a new window] [Download PPT slide]   Figure 1b. Transverse CT images obtained before and after surgery. (a) Two ROIs are shown in the posterior lung parenchyma. This patient subsequently underwent left upper lobectomy. The anterior pair of ROIs are not demonstrated. (b) Image shows postoperative development of ARDS, with two posterior ROIs positioned at an equivalent anatomic location. Again, the anterior pair of ROIs are not demonstrated, since the workstation is able to display only two ROIs at a time.

The expected postoperative density was compared with the observed postoperative density. From these data, increases in density above expected values were compared between the operated and nonoperated lungs by using paired testing. On the side that was subsequently operated on, the preoperative volume was derived from the total volume of that lung minus the volume of the segment or lobe to be resected as measured at the workstation.

Differences between the expected and observed postoperative densities were compared. Because comparisons were made between operated and nonoperated lungs in the same patients, paired tests (ie, the Wilcoxon signed rank test and the paired t test) were used. A P value of less than .05 was considered to indicate a statistically significant difference.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
During the period of review, 583 lobectomies and segmentectomies were performed. Seventeen (2.9%) cases of PLE (15 men; age range, 25–79 years) met the consensus criteria for ARDS. Nine of these patients had undergone pre- and postoperative scanning according to the inclusion criteria detailed in Materials and Methods.

Four patients underwent left upper lobectomy (two of which included the lingula), two underwent left upper lobe division resection, one underwent right upper lobectomy, one underwent left lower lobectomy, and one underwent right upper and middle lobe resection. In all but one case, the underlying abnormality was a bronchogenic primary neoplasm. In the remaining case, the final diagnosis was focal organizing pneumonia.

The interval between surgery and diagnosis of ARDS varied between 1 and 16 days, with a median of 4 days. The interval between diagnosis of ARDS and CT examination was less than 48 hours in eight patients, and less than 72 hours in one patient. Three of the nine patients died following development of ARDS (two on postoperative day 16 and one on postoperative day 35). The remainder survived and were discharged to follow-up.

There was an absolute increase in observed postoperative mean lung density of 79 mg/mL for the operated lung and 376 mg/mL for the nonoperated lung when unadjusted for volume change, and this was apparent from the CT images (Fig 2).

View larger version (141K): [in this window] [in a new window] [Download PPT slide]   Figure 2a. Transverse CT images displayed at lung window settings demonstrate postoperative CT changes in three representative patients. The side of resection was (a) right, (b) left, and (c) right. There is marked asymmetry of parenchymal opacity in each case. The changes—namely, ground glass parenchymal opacification, increased prominence of interlobular septa, and an anteroposterior opacity gradient—are consistent with PLE. The changes are more marked in the nonoperated lung in each case. The small pleural effusions were excluded from the volume measurements. Areas of intense parenchymal opacification (evident posteriorly in a and b) were avoided during ROI placement.

View larger version (144K): [in this window] [in a new window] [Download PPT slide]   Figure 2b. Transverse CT images displayed at lung window settings demonstrate postoperative CT changes in three representative patients. The side of resection was (a) right, (b) left, and (c) right. There is marked asymmetry of parenchymal opacity in each case. The changes—namely, ground glass parenchymal opacification, increased prominence of interlobular septa, and an anteroposterior opacity gradient—are consistent with PLE. The changes are more marked in the nonoperated lung in each case. The small pleural effusions were excluded from the volume measurements. Areas of intense parenchymal opacification (evident posteriorly in a and b) were avoided during ROI placement.

View larger version (155K): [in this window] [in a new window] [Download PPT slide]   Figure 2c. Transverse CT images displayed at lung window settings demonstrate postoperative CT changes in three representative patients. The side of resection was (a) right, (b) left, and (c) right. There is marked asymmetry of parenchymal opacity in each case. The changes—namely, ground glass parenchymal opacification, increased prominence of interlobular septa, and an anteroposterior opacity gradient—are consistent with PLE. The changes are more marked in the nonoperated lung in each case. The small pleural effusions were excluded from the volume measurements. Areas of intense parenchymal opacification (evident posteriorly in a and b) were avoided during ROI placement.

View larger version (22K): [in this window] [in a new window] [Download PPT slide]   Figure 3. Pulmonary CT density change following surgery and development of ARDS. The plot demonstrates the change relative to expected density for the operated and nonoperated lungs in each case. In all patients, there was an increase in density in the nonoperated lung, while in the operated lung, density increased, decreased, or stayed the same. Possible explanations are explored in the Discussion.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
PPE and PLE complicate a small but substantial number of thoracic operations. PPE has been defined in a number of ways, with the development of noncardiogenic pulmonary edema being the common element in all definitions (1–3,6). The course of PPE, in terms of clinical manifestation, physiologic changes, and histopathologic findings, is indistinguishable from that of ARDS (2), and many authorities regard PPE as ARDS in a specific clinical context (13). Because of the varying definitions of PPE and PLE, the exact incidence of each is uncertain. Reported rates of PLE vary from 1% to 15% (2–5,14). PPE of sufficient severity to meet the American-European Consensus Committee on ARDS criteria (15) has been reported in up to 5.2% of patients following lobectomy, whereas acute lung injury, reflected by less severe indices of lung injury, has been reported in 2.2% of cases. Mortality rates for PLE have not been well defined. Estimates of mortality following development of PPE range between 50% and 100% (2–5,15). In our series, the mortality rate was three of nine. Jordan et al (13) reviewed the pathophysiology of PPE, and a number of possible explanations have been explored.

Impaired lymphatic drainage has been suggested as a contributing factor in the development of PPE and PLE (13). However, the relative sparing of the lung subjected to surgery in this series is indirect evidence that impairment of lymphatic drainage is unlikely to be a precipitating cause. Similarly, the suggestion that the surgical manipulation necessary to carry out hilar or mediastinal dissection might precipitate the development of an inflammatory reaction (13) also appears to be an unlikely explanation, given the predominantly contralateral lung involvement shown in our study. It is possible, however, that inflammatory mediators released from the manipulated lung may somehow produce preferential contralateral lung injury (13).

Another potential source of inflammatory mediators is transfusion products, particularly fresh frozen plasma (FFP) (5). Van der Werff et al (5) found that patients who developed PPE had been administered substantially larger amounts of FFP than that administered to a control group. They suggest that one possible mechanism to explain this finding might be the presence of stable and unstable components of blood coagulation, fibrinolytic, and complement systems in FFP. In addition, they postulate that antibodies in FFP may activate leukocytes and granulocytes within the recipient and give rise to capillary leakage and pulmonary edema. They draw a parallel with a syndrome known to occur occasionally after administration of FFP in nonpneumonectomy patients, known as transfusion-related acute lung injury, or TRALI (5).

The technique of one-lung ventilation is performed routinely in most pulmonary resections. During surgery, the contralateral lung remains ventilated and provides the sole means of adequate oxygenation for the patient. Although a substantial proportion of cardiac output will be diverted to the contralateral lung, the remaining proportion of cardiac output that passes through the nonventilated lung will result in a physiologic shunt and reduce oxygenation, which can be counteracted by increasing the inspiratory oxygen concentration. Consequently, the single ventilated lung is subjected to a combination of increased alveolar oxygen tensions, increased flow, and hyperinflation.

The threshold at which oxygen toxicity causes lung damage is not accurately defined, but there is extensive evidence that hyperoxia per se produces oxidative stress and diffuse alveolar damage (16). Barotrauma from mechanical ventilation may also occur, and during single-lung ventilation, peak respiratory pressures of up to 30 cm of H₂O and high tidal volumes of 10 mL/kg are customarily used (17). If there is preexisting disease or reduced compliance, as might be expected in patients who develop bronchogenic carcinoma (possibly as a result of a lifetime of smoking), it may be necessary to apply higher tidal volume or peak respiratory pressures (17). Furthermore, ipsilateral mediastinal shift following lobectomy may occur, resulting in an increase in the functional residual capacity of the nonresected lung (18). Single lung ventilation with subsequent hyperinflation may therefore result in lung trauma that preferentially affects the nonoperated side.

The possibility that ischemic reperfusion injury occurs in the operated lung has also been raised as a possible precipitating mechanism (13). During surgery, the operated lung is excluded from the ventilatory circuit and collapsed to facilitate surgical dissection of the hilum and mediastinum. In this way, it is dependent on the bronchial circulation and has the potential to become relatively ischemic. When lobectomy is performed, the nonresected portion of the operated lung is then re-expanded and reperfused. The cycle of collapse and re-expansion may be repeated several times during the course of lobe resection. This cycle of ischemia and reperfusion may release inflammatory mediators into the circulation (13). While the lung or lobe on the side of operation is collapsed, there will be an increase in blood flow to the contralateral lung. This in turn will increase the shear stress on the pulmonary vascular endothelium, particularly in patients with chronic obstructive pulmonary disease and reduced pulmonary vascular compliance (13). The collapsed lung will be relatively protected from potential shear stress injury, and this mechanism, possibly exacerbated by the release of inflammatory mediators, may contribute to or account for the asymmetric pattern of disease.

Examination of density changes in this series (eg, Figs 2 and 3) demonstrated that density in the nonoperated lung increased in all cases, whereas density in the operated lung increased in four cases, stayed essentially the same in three cases, and decreased in two cases. This range of change suggests that whatever the mechanism or combination of mechanisms might be, the involvement of the operated lung shows marked variability. In the two cases where there was a reduction in density in the operated lung, it seems that there must be a reduction in the tissue water content between the preoperative and postoperative parenchyma, since allowance was made for changes in lung volume. A vasospastic mechanism would seem a likely explanation for this change, and this may be an element of the relative sparing effect observed in the other cases, although masked by other unexplained influences.

The possibility exists that our data underestimate the magnitude of the observed discrepancy in lung density between the operated and nonoperated lungs. Intense parenchymal opacification was more extensive in the nonoperated lung. Areas of intense parenchymal opacification would be expected to result in less aerated lung in the remainder of the hemithorax. This reduced amount of aerated lung might also be expected to undergo compensatory overinflation and counteract the increases in density as a result of PLE. A further potential difficulty is the inclusion of areas of dense parenchymal opacification (in the postoperative lung) or tumor masses (in the preoperative lung) in global density measurements. Therefore, we chose to measure lung density by using ROIs rather than by calculating the density of the entire lung volume. The random variation or noise in the density measurement generated by manually selecting ROIs might be expected to reduce, rather than accentuate, the differences between the two sides.

We recognize that attenuation measurements will be altered by the administration of intravenous contrast material during the preoperative examination. However, the effects of contrast enhancement should be equal for each lung. Furthermore, contrast material administration would not accentuate but would rather reduce the magnitude of the general increase in measured attenuation detected in the lungs following development of PLE, since contrast material was not administered postoperatively. There is a difference in ROI volumes between preoperative and postoperative examinations due to differences in collimation. Although this might slightly influence absolute density measurement (19), it should not alter the relative differences between the right and left lungs at preoperative and postoperative examination.

Thus, a number of mechanisms proposed for modulation of lung injury following pneumonectomy may have different effects on the operated and nonoperated sides when lobectomy is performed, including ischemia reperfusion, high oxygen tensions, high inflation pressures, and alterations in lymphatic drainage. The relative importance of the various mechanisms at work has yet to be fully defined, but the demonstration of marked asymmetry in lung damage following lobectomy has been confirmed objectively by the results of this study.

	FOOTNOTES

 
Abbreviations: ARDS = acute respiratory distress syndrome, FFP = fresh frozen plasma, PLE = postlobectomy edema, PPE = postpneumonectomy pulmonary edema, ROI = region of interest

Author contributions: Guarantor of integrity of entire study, S.P.G.P.; study concepts and design, S.P.G.P., D.M.H., A.U.W.; literature research, S.P.G.P.; clinical studies, S.P.G.P., S.J.J.; data acquisition, S.P.G.P.; data analysis/interpretation, S.P.G.P., D.M.H., A.U.W.; statistical analysis, A.U.W.; manuscript preparation and definition of intellectual content, S.P.G.P.; manuscript editing, revision/review, and final version approval, all authors.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Zeldin RA, Normandin D, Landtwing D, Peters RM. Postpneumonectomy pulmonary edema. J Thorac Cardiovasc Surg 1984; 87:359-365.
2. Turnage WS, Lunn JJ. Postpneumonectomy pulmonary edema: a retrospective analysis of associated variables. Chest 1993; 103:1646-1650.
3. Waller DA, Gebitekin C, Saunders NR, Walker DR. Noncardiogenic pulmonary edema complicating lung resection. Ann Thorac Surg 1993; 55:140-143.
4. Mathru M, Blakeman B, Dries DJ, Kleinman B, Kumar P. Permeability pulmonary edema following lung resection. Chest 1990; 98:1216-1218.
5. van der Werff YD, van der Houwen HK, Heijmans PJ, et al. Postpneumonectomy pulmonary edema: a retrospective analysis of incidence and possible risk factors. Chest 1997; 111:1278-1284.
6. Shapira OM, Shahian DM. Postpneumonectomy pulmonary edema. Ann Thorac Surg 1993; 56:190-195.
7. Bernard GR, Artigas A, Brigham KL, et al. The American-European Consensus Conference on ARDS: definitions, mechanisms, relevant outcomes, and clinical trial coordination. Am J Respir Crit Care Med 1994; 149:818-824.
8. Hedlund LW, Vock P, Effmann EL. Computed tomography of the lung: densitometric studies. Radiol Clin North Am 1983; 21:775-788.
9. Millar AB, Denison DM. Vertical gradients of lung density in healthy supine men. Thorax 1989; 44:485-490.
10. Robinson PJ, Kreel L. Pulmonary tissue attenuation with computed tomography: comparison of inspiration and expiration scans. J Comput Assist Tomogr 1979; 3:740-748.
11. Verschakelen JA, Van Fraeyenhoven L, Laureys G, Demedts M, Baert AL. Differences in CT density between dependent and nondependent portions of the lung: influence of lung volume. AJR Am J Roentgenol 1993; 161:713-717.
12. Collins CD, Wells AU, Hansell DM, et al. Observer variation in pattern type and extent of disease in fibrosing alveolitis on thin section computed tomography and chest radiography. Clin Radiol 1994; 49:236-240.
13. Jordan S, Mitchell JA, Quinlan GJ, Goldstraw P, Evans TW. The pathogenesis of lung injury following pulmonary resection. Eur Respir J 2000; 15:790-799.
14. Parquin F, Marchal M, Mehiri S, Herve P, Lescot B. Post-pneumonectomy pulmonary edema: analysis and risk factors. Eur J Cardiothorac Surg 1996; 10:929-932.
15. Hayes JP, Williams EA, Goldstraw P, Evans TW. Lung injury in patients following thoracotomy. Thorax 1995; 50:990-991.
16. Ichikado K, Suga M, Gushima Y, et al. Hyperoxia-induced diffuse alveolar damage in pigs: correlation between thin-section CT and histopathologic findings. Radiology 2000; 216:531-538.
17. Gothard JW. High frequency jet ventilation during bronchial surgery (letter). Anaesthesia 1988; 43:511-512.
18. Larsson A, Malmkvist G, Werner O. Variations in lung volume and compliance during pulmonary surgery. Br J Anaesth 1987; 59:585-591.
19. Kemerink GJ, Kruize HH, Lamers RJ, van Engelshoven JM. CT lung densitometry: dependence of the CT number histograms on the sample volume and consequences for scan protocol comparability. J Comput Assist Tomogr 1997; 21:948-954.

This article has been cited by other articles:

				A. Kozian, T. Schilling, H. Schutze, F. Heres, T. Hachenberg, and G. Hedenstierna Lung computed tomography density distribution in a porcine model of one-lung ventilation Br. J. Anaesth., April 1, 2009; 102(4): 551 - 560. [Abstract] [Full Text] [PDF]

				A. L. Lagan, G. J. Quinlan, S. Mumby, D. D. Melley, P. Goldstraw, G. J. Bellingan, M. R. Hill, D. Briggs, P. Pantelidis, R. M. du Bois, et al. Variation in Iron Homeostasis Genes Between Patients With ARDS and Healthy Control Subjects Chest, June 1, 2008; 133(6): 1302 - 1311. [Abstract] [Full Text] [PDF]

				N. Alam, B. J. Park, A. Wilton, V. E. Seshan, M. S. Bains, R. J. Downey, R. M. Flores, N. Rizk, V. W. Rusch, and D. Amar Incidence and Risk Factors for Lung Injury After Lung Cancer Resection Ann. Thorac. Surg., October 1, 2007; 84(4): 1085 - 1091. [Abstract] [Full Text] [PDF]

				T. Schilling, A. Kozian, M. Kretzschmar, C. Huth, T. Welte, F. Buhling, G. Hedenstierna, and T. Hachenberg Effects of propofol and desflurane anaesthesia on the alveolar inflammatory response to one-lung ventilation Br. J. Anaesth., September 1, 2007; 99(3): 368 - 375. [Abstract] [Full Text] [PDF]

				E. J. Chae, J. B. Seo, S. Y. Kim, K.-H. Do, J.-N. Heo, J. S. Lee, K. S. Song, J. W. Song, and T.-H. Lim Radiographic and CT Findings of Thoracic Complications after Pneumonectomy RadioGraphics, September 1, 2006; 26(5): 1449 - 1468. [Abstract] [Full Text] [PDF]

				K. Ichikado, M. Suga, H. Muranaka, Y. Gushima, H. Miyakawa, M. Tsubamoto, T. Johkoh, N. Hirata, T. Yoshinaga, Y. Kinoshita, et al. Prediction of Prognosis for Acute Respiratory Distress Syndrome with Thin-Section CT: Validation in 44 Cases Radiology, December 1, 2005; 238(1): 321 - 329. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) All Versions of this Article: 2232010721v1 223/2/468    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Padley, S. P. G. Articles by Hansell, D. M. Search for Related Content PubMed PubMed Citation Articles by Padley, S. P. G. Articles by Hansell, D. M.