Air Trapping on CT of Patients with Pulmonary Embolism

Transcription

1 Hiroaki rakawa 1,2 Yasuyuki Kurihara 1 Kaoru Sasaka 1 Yasuo Nakajima 1 W. Richard Webb 3 Received ugust 3, 2001; accepted after revision November 14, Department of Radiology, St. Marianna University School of Medicine, Sugao, Miyamae-Ku, Kawasaki ity, Japan. 2 Present address: Department of Radiology, Dokkyo University School of Medicine, Mibu, Shimotsuga-gun, Tochigi, Japan. ddress correspondence to H. rakawa. 3 Department of Radiology, University of alifornia San Francisco, 505 Parnassus ve., San Francisco, JR 2002;178: X/02/ merican Roentgen Ray Society ir Trapping on T of Patients with Pulmonary Embolism OJETIVE. We evaluated the relationship of air trapping to mosaic perfusion in patients with pulmonary embolism. SUJETS ND METHODS. Forty-one consecutive patients with suspected pulmonary embolism underwent expiratory T followed by helical T angiography. fter excluding 12 patients who had airway disease or were smokers, we divided the patients into two groups: those with (n = 15) and without (n = 14) pulmonary embolism. For each patient, six expiratory images were evaluated for the presence of air trapping, and the corresponding six images from T angiography were evaluated for the presence of mosaic perfusion. lot locations were assessed on T angiography and were correlated with the presence of air trapping and mosaic perfusion. RESULTS. In patients with pulmonary embolism, mosaic perfusion was identified in 32 areas (seven patients, 46.7%), and air trapping was identified 68 areas (nine patients, 60%). Of the 32 areas of mosaic perfusion, 23 areas (71.9%) showed air trapping on expiratory T scans. Of the 68 areas with air trapping on expiratory scans, 23 areas (33.8%) showed mosaic perfusion on inspiratory scans, and 44 areas (64.7%) had clots in the arteries leading to them. lots were more frequently identified in areas of lower attenuation on inspiratory T scans and air trapping (21/ 23) than in those of normal attenuation on inspiratory T scans and air trapping (23/45) (p < 0.005). Only one patient without pulmonary embolism had air trapping (p < 0.005). ONLUSION. ir trapping is common in pulmonary embolism and may be the cause of mosaic perfusion. ir trapping can be seen distal to vessels not showing pulmonary embolism. I nhomogeneous lung attenuation seen on high-resolution T may reflect regional differences in pulmonary blood volume, a finding termed mosaic perfusion [1, 2]. Mosaic perfusion may be present in patients who have airway obstruction with reflex vasoconstriction [3 9] or in patients with vascular obstruction such as that occurring in pulmonary embolism [10]. In many patients, a distinction may be made between airway and vascular obstruction as a cause of mosaic perfusion using expiratory scans [11]: ir trapping is commonly present in patients with mosaic perfusion related to airway disease, but it is said to be absent in patients with pulmonary embolism. However, it has been suggested that air trapping also may be seen on expiratory high-resolution T in patients with pulmonary embolism [12]. The objective of our study was to determine the frequency of air trapping in patients with pulmonary embolism and to evaluate the relationship of air trapping to abnormalities in lung attenuation seen on inspiratory scans. Subjects and Methods Patients Forty-one consecutive patients referred for suspected pulmonary embolism were included in our study. Inclusion criteria were high suspicion of acute or chronic pulmonary embolism and a stable cardiac and respiratory status. High suspicion of pulmonary embolism was based on the results of clinical history, physical examination, EG, echocardiography, venography, or chest radiography. Exclusion criteria were age younger than 18 years, pregnancy, renal insufficiency (serum creatinine level > 1.7 mg/dl), history of anaphylactic reaction to contrast material, or unstable respiratory condition. We obtained informed consent from all patients before performing chest T. Patients with suspected pulmonary embolism underwent chest T as the initial screening test, followed by ventilation perfusion radionuclide scanning, and, in selected cases, pulmonary angiography. The diagnosis of pulmonary embolism was made in 23 patients on the basis of high-probability JR:178, May

2 rakawa et al. ventilation perfusion radionuclide scans (n = 18) or abnormal findings on pulmonary angiography (n = 5). Eight of the patients with pulmonary embolism were excluded because they had a history of pulmonary parenchymal disease or were smokers. Thus, our study included 15 patients with pulmonary embolism who neither were smokers nor had a history of pulmonary disease (seven men and eight women; age range, years; mean age, 52.7 years). In these 15 patients, the diagnosis of pulmonary embolism was established on the basis of high-probability ventilation perfusion radionuclide scanning (n = 11) or abnormal findings on pulmonary angiography (n = 4). Five patients had deep vein thrombosis confirmed by venography of the lower extremities; one patient had vasculitis; one patient had protein deficiency; and in the remaining eight patients, the pulmonary embolism had no identifiable cause. Nine patients were considered to have acute pulmonary embolism, two had chronic pulmonary embolism, and four had acute and chronic pulmonary embolism. ngiographic criteria for the diagnosis of acute pulmonary embolism included complete or partial filling defects as the sole abnormalities or as observed in association with arterial cutoffs, perfusion defects, or both [13]. On T, emboli were considered acute if they completely or partially occluded the arterial lumen. The diagnosis of chronic pulmonary embolism was suggested by historic and clinical features and was confirmed by combined results from perfusion lung scanning, pulmonary angiography, echocardiography, and or helical T angiography. ngiographic criteria for the diagnosis of chronic pulmonary embolism were pouching defects, webs or bands, intimal irregularities, abrupt vascular narrowing, compete vascular obstruction, or any combination of these findings [14]. On T, the diagnosis of chronic pulmonary embolism required at least one of the following: emboli eccentric and contiguous with the vessel wall, evidence of recanalization, arterial stenosis or web, or reduction of more than 50% of the overall arterial diameter [15]. cute pulmonary embolism in patients with chronic pulmonary embolism was suspected on the basis of clinical features of an acute episode of chest pain, dyspnea, or hemoptysis in patients with chronic pulmonary embolism and was confirmed by the presence of acute clots on T (n = 3) or newly developed perfusion defects on perfusion scanning (n = 1). Eighteen of the 41 patients studied were considered not to have pulmonary embolism because of concordant negative findings on helical T angiography and ventilation perfusion scintigraphy and negative findings at clinical follow-up of at least one month. Of the 18 patients without pulmonary embolism, 14 patients had no known airway disease, and they served as a control group (seven men and seven women; age range,19 67 years; mean age, 49.5 years). The diagnostic workup was performed by one of the authors and another staff radiologist at our institution. Discordance was resolved by the opinion of a third staff radiologist. T Technique Helical T angiography was performed from above the aortic arch to the diaphragm at deep inspiration during a single breath-hold. T scans were obtained in all patients using X-Vigor or X-Vision T scanners (Toshiba, Tokyo, Japan). bolus of 100 ml of iodinated contrast material (iopamidol 61%, Iopamiron; Nihon Schering, Osaka, Japan) was injected through the antecubital vein at a rate of 3 ml/sec with a scan delay of 15 sec. Scanning parameters included a 3-mm collimation, a 5-mm/sec table speed, and reconstruction at 3- or 1.5-mm intervals. We also obtained expiratory high-resolution T scans at the time of helical T pulmonary angiography. Expiratory high-resolution T was performed at deep exhalation (expiratory scans) using a 2-mm collimation. Scans were obtained at six evenly spaced levels from lung apices to bases. In general, scans were obtained at the following levels: above the aortic arch, between the aortic arch and the tracheal carina, at the tracheal carina, between the tracheal carina and the confluence of pulmonary veins, at the confluence of pulmonary veins, and near the diaphragm. oth mediastinal window images (window width, H; level, H) and lung window images (window width, 1200 H; level, 750 H) were obtained. Lung window scans were reconstructed using a high-spatial-frequency algorithm. Image nalysis T scans were interpreted by consensus of two radiologists. In the first session, reviewing both the hard copy and the monitor image, clot location was recorded in the main, lobar, or segmental arteries in the right or left lung. When obvious arterial changes suggestive of chronic embolism (e.g., arterial stenosis or web, reduction of > 50% of the overall arterial diameter) were identified, we recorded the artery as having a clot even if a low-attenuation embolus was not clearly identified. If multiple clots were identified in both the proximal and distal portions of the artery, only the clot in the proximal portion was recorded. For example, when a clot was identified in the right interlobar artery, those clots in the lobar and segmental arteries of right lower and middle lobes were not recorded. Identification of segmental arteries on T images required an analysis of both mediastinal and lung window settings; bronchi were used to accurately identify specific segmental arteries. Opacification of pulmonary arteries was analyzed at mediastinal window settings. While viewing on a workstation, we changed lung window widths and levels when contrast material in the pulmonary artery was considered sufficiently dense to mask a clot. In the second session, lung parenchyma was evaluated on hard copy that was photographed with lung window settings, and attention was paid to any lung attenuation abnormalities. Lung window images were reviewed in a session separate from the one dedicated to searching for the clot; the interval between the two sessions was 6 months. We selected inspiratory scans on lung windows that corresponded to the same levels as those of expiratory high-resolution scans, and the presence and location of mosaic perfusion were recorded as defined by areas of decreased lung attenuation with or without reduced vessel size in a lobe [16]. ecause we obtained six expiratory slices of each patient, only six inspiratory levels corresponding to the six expiratory levels were selected. When reviewing expiratory scans, we recorded the presence and location of air trapping. ir trapping was considered present if areas of lung showed a less than normal increase in attenuation after expiration or showed little change in the cross-sectional area [10]. ecause air trapping can be seen in isolated secondary pulmonary lobules even in healthy subjects [17], we excluded areas of air trapping seen in a single secondary pulmonary lobule. One area each of mosaic perfusion and air trapping was determined as an area of homogeneously lower attenuation in a lobe on inspiratory and expiratory images, respectively. The extent and location of mosaic perfusion on an inspiratory scan and air trapping on the corresponding expiratory scan were compared and were related to the presence and location of the clot. If one area of air trapping on an expiratory scan involved more than two areas of mosaic perfusion on the corresponding inspiratory scan, the area of air trapping was divided and counted as the same number of areas of mosaic perfusion. Statistical nalysis The difference in prevalence of air trapping between patients with pulmonary embolism and patients without pulmonary embolism was assessed using the chi-square test. mong the areas with air trapping, the difference in the prevalence of clots was assessed between the areas of lower attenuation and those of normal attenuation on the corresponding inspiratory scan levels using the chi-square test. We considered a p value greater than 0.05 a statistically significant difference. Results In the first session, helical T angiography showed clots in 14 patients (eight with acute pulmonary embolism, two with chronic pulmonary embolism, and four with acute and chronic pulmonary embolism). In one patient with acute pulmonary embolism, a clot was not identified on T; the diagnosis was based on the highprobability findings on ventilation perfusion scans. lots were identified in four main, 20 lobar, and 25 segmental arteries. In one patient with acute pulmonary embolism, a clot was not identified on helical T angiography. In the second session, seven expiratory slices were excluded because of poor exhalation (four slices in two patients) or mismatch between inspiratory and expiratory scan levels (three slices in two patients). mong 15 patients with pulmonary embolism, nine patients (60.0%) were considered to have air trapping (four with acute pulmonary embolism, one 1202 JR:178, May 2002

3 ir Trapping on T of Patients with Pulmonary Embolism with chronic pulmonary embolism, and four with acute and chronic pulmonary embolism). In distinction, among 14 patients without pulmonary embolism, only one patient showed air trapping ( p < 0.005). ir trapping was identified in 68 areas from nine patients with pulmonary embolism, and mosaic perfusion was identified in 32 areas from seven patients (four with acute pulmonary embolism and three with acute and chronic pulmonary embolism). These seven patients also had air trapping. reas of air trapping were identified in the right upper lobe (n = 18), left upper lobe (n = 13), right middle lobe (n = 4), right lower lobe (n = 19), and left lower lobe (n = 14). reas of mosaic perfusion were identified in the right upper lobe (n = 5), left upper lobe (n = 10), right middle lobe (n = 3), right lower lobe (n = 9), and left lower lobe (n = 5). Thirty-two areas showed decreased lung attenuation on the inspiratory scan, and 20 of these areas (62.5%) showed reduced vessel size as a sign of mosaic perfusion (Fig. 1). lots were identified on helical T angiography in 29 arteries (90.6%) proximal to the areas of decreased attenuation. Of the 32 areas of decreased attenuation on the inspiratory scans, 23 areas (71.9%) in seven patients showed air trapping on the expiratory scans. Of the nine areas with decreased attenuation not associated with air trapping, eight (88.9%) of the areas had clots in the proximal arteries. In 68 areas with air trapping, clots were identified on helical T angiography in 44 arteries (64.7%) proximal to those areas (Figs. 2 and 3). However, in the remaining 24 areas, no clot was identified (Fig. 4). Forty-five areas (66.2%) showed normal attenuation on the inspiratory scans, and 23 of these areas (51.1%) showed clots in the arteries proximal to them. Twenty-three (33.8%) of the 68 areas that showed air trapping on the expiratory scans showed decreased attenuation on the inspiratory scans, and 21 of these areas (91.3%) had clots in the arteries proximal to them. The prevalence of clots in the proximal artery was significantly higher in areas of air trapping with mosaic perfusion (mosaic perfusion due to air trapping) Fig. 1. cute pulmonary embolism in 76-year-old woman., Helical T angiogram (3-mm collimation) shows clot (arrowhead) in posterior segmental arteries of right upper lobe. lot was not identified in arteries of left upper lobe., Helical T angiogram obtained using lung window settings shows inhomogeneous lung attenuation (arrows) in both upper lobes. Vessel size in lower attenuation areas is only slightly reduced, thus indicating mosaic perfusion as possible cause of inhomogeneous lung attenuation., Expiratory high-resolution T scan obtained at same level as shows air trapping to be associated with mosaic perfusion in both upper lobes. ir trapping in left upper lobe closely corresponds to area of low attenuation seen on inspiratory scan (). In right upper lobe, area of air trapping appears larger than area of mosaic perfusion seen on inspiratory scan. JR:178, May

4 rakawa et al. than in areas of air trapping with normal lung attenuation on inspiratory scanning (p < 0.005). Discussion On T, reduced lung attenuation or mosaic perfusion has been reported to occur in 7 12% of patients with acute pulmonary embolism and in more than 70% of those with chronic pulmonary embolism [18 21]. direct reduction in pulmonary blood volume as a result of arterial occlusion is usually considered to be responsible for lower lung attenuation in those patients [22]. However, Im et al. [23] reported that they did not find lung attenuation changes in pigs during the 28 days immediately after experimentally induced pulmonary embolism using radiopaque silicon spheres. These researchers raised concerns that acute reduction of pulmonary blood volume may not affect lung attenuation in acute pulmonary embolism. t the same time, they confirmed that bronchial obstruction in the pigs resulted in lobular areas of low attenuation on high-resolution T. Kim et al. [24] reported similar findings in experimentally induced pulmonary embolism in pigs. In their study, thin-section T scans were obtained as long as 12 weeks after pulmonary artery obstruction with a detachable balloon in seven pigs. They found no statistically significant difference between the measured lung attenuation of the normal lung and that of the lung distal to the obstruction. It has been suggested that air trapping may be seen on expiratory high-resolution T in patients with pulmonary embolism, and that it is responsible for the areas of low attenuation [12]. Worthy et al. [12] reported that two of four patients with vascular disease in their study showed air trapping on expiratory scans. sthmalike wheezing is reported to occur in patients with acute pulmonary embolism and is attributed to bronchoconstriction in those patients [25 28]. Initially, Gurewich et al. [26] reported seven patients with acute pulmonary embolism who showed evidence of bronchoconstriction on spirometry. One of their patients presented with expiratory wheezing. This bronchoconstriction was relieved by IV administration of heparin in four patients. In a subsequent larger series, Sasahara et al. [25] studied 72 patients with acute pulmonary embolism and confirmed the presence of bronchoconstriction on spirometry in most of their patients. The bronchoconstriction in their series also showed improvement after IV administration of a bolus of heparin. The proposed mechanism of bronchoconstriction in acute pulmonary embolism includes bronchoactive amines such as serotonin and prostaglandins that are released from platelet aggregations in the thrombus, as has been shown in canine lungs Fig. 2. cute embolism in 68-year-old woman with deep vein thrombosis., Helical T angiogram shows massive thromboembolism in right descending pulmonary artery and its branches in lower lobes. lot (arrow) is also visualized in artery leading to left lingula. lot was also identified in segmental arteries in left lower lobe (not shown)., Helical T angiogram obtained using lung window settings shows inhomogeneous lung attenuation with peripheral ground-glass opacity (arrows) in both lower lobes., Expiratory high-resolution T scan obtained at same level as shows multifocal areas of air trapping distal to arteries with clots JR:178, May 2002

5 ir Trapping on T of Patients with Pulmonary Embolism [29], or a change in parasympathetic nervous system tension, which controls the bronchial smooth-muscle tension [30]. Our study, using expiratory T, showed that air trapping is not rare in patients with pulmonary embolism and that air trapping was associated with areas of mosaic perfusion in 71.9% of the patients we studied. In these areas, the cause of mosaic perfusion was attributed to air trapping. ir trapping on expiratory T was seen not only in areas with pulmonary embolism (64.7%) but also in areas without embolism (35.3%). In a report of experimentally induced unilateral pulmonary embolism in dogs, ustin and Sagel [30], using 24-hr-old human venous thrombi as embolic material, observed that bronchial narrowing did not differ quantitatively between the ipsilateral and contralateral lungs. They concluded that bronchial narrowing is a generalized response, mediated by humoral factors or the parasympathetic nervous system. Our results also support this conclusion. Furthermore, our study showed that the presence of a clot was more frequently associated with areas of mosaic perfusion due to air trapping than with areas of normal attenuation with air trapping on inspiratory scans. ecause lung attenuation is determined by the amount of air in the alveoli, pulmonary interstitium, and pulmonary blood volume, lower attenuation in areas distal to the artery with pulmonary embolism can be attributed to increased air content, reduced blood volume, or both. We think that bronchoconstriction might be more severe in areas with pulmonary embolism than in areas without pulmonary embolism. It is possible that both the direct reduction of pulmonary blood volume by a clot and air trapping contribute to the lower lung attenuation on inspiratory T, although findings in the experimental studies of Im et al. [23] and Kim et al. [24] suggest that this is not likely to be the case. In acute pulmonary embolism, the presence of mosaic perfusion is considered uncommon. oche et al. [21] compared lung parenchymal findings of 88 patients with suspected acute pulmonary embolism who underwent helical T pulmonary angiography. Those researchers found mosaic perfusion in 12% of patients with acute pulmonary embolism and in 10% of patients without this finding. Shah et al. [18] evaluated parenchymal findings in 28 patients with acute pulmonary embolism and found only 7% with mosaic perfusion. However, in our series, Fig. 3. cute embolism in 61-year-old woman with chronic pulmonary embolism., Helical T angiogram shows eccentric clots (arrows) in both lower lobe pulmonary arteries., Helical T angiogram obtained using lung window settings shows near normal lung attenuation. Focal pleural-based ground-glass attenuation (arrow) is visible in right lower lobe., Expiratory high-resolution T scan obtained at same level as shows extensive areas of air trapping (arrows) distal to arteries with clots. JR:178, May

6 rakawa et al. we found mosaic perfusion in four (44.4%) of nine patients with acute pulmonary embolism a higher frequency than expected. This high frequency in our series may result from a variability of patient selection and our special attention to changes in lung attenuation. In our series, air trapping was observed in a patient with chronic pulmonary embolism despite the fact that air trapping is rarely reported in the literature [12]. Im et al. [23] reported that they found a patient with air trapping in a chronic pulmonary embolism who had Takayasu s arteritis. Remy-Jardin et al. [31] reported progressive cylindric dilatation of bronchi in 21 of 33 patients with chronic pulmonary embolism who were followed up with T, and they suggested possible airway changes in this disease. Pulmonary function test results in their series showed a normal range of mean forced expiratory volume in 1 sec, but they found that the mean value of the maximum expiratory flow between 25% and 75% of the forced vital capacity was reduced. They speculated that morphologic and functional derangements occurred in the small airways in their patients as a result of chronically absent pulmonary arterial perfusion [31]. We think that air trapping in our patient with chronic pulmonary embolism is likely the result of airway changes caused by chronic pulmonary artery obstruction, as suggested by Remy-Jardin et al., although our patient had no identifiable bronchial dilatation. nother possible mechanism of air trapping in patients with chronic pulmonary embolism may be repeated undiagnosed acute embolization. Our study has several limitations. First, helical T angiography was used in most of the patients to determine the presence of clots and their location. The diagnostic accuracy of helical T angiography is comparable to that of conventional angiography in the diagnosis of both acute and chronic pulmonary embolism involving the proximal arteries [32 34]. However, small clots in the peripheral arteries may be missed on helical T angiography [35]. Second, the inspiratory scans used for comparison in most patients were obtained after injection of contrast material. s a result, we could not measure lung density differences between inspiratory scans and expiratory scans, and our observation of air trapping was based solely on the subjective visual assessment of lung attenuation. Only those areas showing obvious low attenuation or no change of cross-sectional area after expiration were considered to show air trapping. Third, our study included a relatively small number of patients. In conclusion, air trapping occurred in 60% of patients in our study with the T finding of pulmonary embolism. ir trapping was associated with mosaic perfusion in 71.9% of lung regions and occurred not only in areas distal to visible clots but also in areas that were well perfused. In an appropriate clinical setting, pulmonary embolism should be considered in the differential diagnosis of diseases associated with mosaic perfusion and air trapping. Fig. 4. cute pulmonary embolism in 45-year-old woman., Helical T angiogram obtained at level between aortic arch and tracheal carina shows clot (arrowhead) in segmental artery of right upper lobe, but no clot was identified in left side., Helical T angiogram obtained above aortic arch using lung window settings shows small area of lower attenuation (arrows) in left upper lobe., Expiratory high-resolution T scan obtained at same level as shows extensive areas of air trapping in both upper lobes (arrows). reas of air trapping are more widespread than areas of mosaic perfusion shown on JR:178, May 2002

7 ir Trapping on T of Patients with Pulmonary Embolism References 1. rakawa H, Webb WR. ir trapping on expiratory high-resolution T scans in the absence of inspiratory scan abnormalities: correlation with pulmonary function tests and differential diagnosis. JR 1998;170: Webb WR, Muller NL, Naidich DP. Standardized terms for high-resolution computed tomography of the lung: a proposed glossary. J Thorac Imaging 1993;8: Franquet T, Diaz, Domingo P, Gimenez, Geli. ir trapping in primary Sjogren syndrome: correlation of expiratory T with pulmonary function tests. J omput ssist Tomogr 1999;23: Hansell DM, Wells U, Rubens M, ole PJ. ronchiectasis: functional significance of areas of decreased attenuation at expiratory T. Radiology 1994;193: Hansell DM, Rubens M, Padley SP, Wells U. Obliterative bronchiolitis: individual T signs of small airways disease and functional correlation. Radiology 1997;203: Hansell DM, Milne DG, Wilsher ML, Wells U. Pulmonary sarcoidosis: morphologic associations of airflow obstruction at thin-section T. Radiology 1998;209: Verschakelen J, Scheinbaum K, ogaert J, Demedts M, Lacquet LL, aert L. Expiratory T in cigarette smokers: correlation between areas of decreased lung attenuation, pulmonary function tests and smoking history. Eur Radiol 1998;8: Newman K, Lynch D, Newman LS, Ellegood D, Newell JD Jr. Quantitative computed tomography detects air trapping due to asthma. hest 1994;106: Small JH, Flower D, Traill Z, Gleeson FV. irtrapping in extrinsic allergic alveolitis on computed tomography. lin Radiol 1996;51: Stern EJ, Muller NL, Swensen SJ, Hartman TE. T mosaic pattern of lung attenuation: etiologies and terminology. J Thorac Imaging 1995;10: rakawa H, Webb WR, Mcowin M, Katsou G, Lee KN, Seitz RF. Inhomogeneous lung attenuation at thin-section T: diagnostic value of expiratory scans. Radiology 1998;206: Worthy S, Muller NL, Hartman TE, Swensen SJ, Padley SP, Hansell DM. Mosaic attenuation pattern on thin-section T scans of the lung: differentiation among infiltrative lung, airway, and vascular diseases as a cause. Radiology 1997;205: Sagel SS, Greenspan RH. Nonuniform pulmonary arterial perfusion: pulmonary embolism? Radiology 1971;99: uger W, Fedullo P, Moser K, uchbinder M, Peterson KL. hronic major-vessel thromboembolic pulmonary artery obstruction: appearance at angiography. Radiology 1992;182: Remy-Jardin M, Remy J, Deschildre F, et al. Diagnosis of pulmonary embolism with spiral T: comparison with pulmonary angiography and scintigraphy. Radiology 1996;200: Remy-Jardin M, Remy J. Spiral T angiography of the pulmonary circulation. Radiology 1999;212: Webb WR, Stern EJ, Kanth N, Gamsu G. Dynamic pulmonary T: findings in healthy adult men. Radiology 1993;186: Shah, Davis SD, Gamsu G, Intriere L. Parenchymal and pleural findings in patients with and patients without acute pulmonary embolism detected at spiral T. Radiology 1999;211: Schwickert H, Schweden F, Schild HH, et al. Pulmonary arteries and lung parenchyma in chronic pulmonary embolism: preoperative and postoperative T findings. Radiology 1994;191: ergin J, Rios G, King M, elezzuoli E, Luna J, uger WR. ccuracy of high-resolution T in identifying chronic pulmonary thromboembolic disease. JR 1996;166: oche EE, Muller NL, Kim KI, Wiggs R, Mayo JR. cute pulmonary embolism: ancillary findings at spiral T. Radiology 1998;207: King M, ergin J, Yeung DW, et al. hronic pulmonary thromboembolism: detection of regional hypoperfusion with T. Radiology 1994;191: Im JG, hoi YW, Kim HD, Jeong YK, Han M. Thin-section T findings of the lungs: experimentally induced bronchial and pulmonary artery obstruction in pigs. JR 1996;167: Kim TK, Im JG, Kim SH, et al. Experimentally induced pulmonary arterial occlusion with detachable balloon in pigs: thin-section T findings. cad Radiol 1998;5: Sasahara, annilla JE, Morse RL, Sidd JJ, Tremblay GM. linical and physiologic studies in pulmonary thromboembolism. m J ardiol 1967;20: Gurewich V, Thomas D, Stein M, Wessler S. ronchoconstriction in the presence of pulmonary embolism. irculation 1963;27: Webster JR Jr, Saadeh G, Eggum PR, Suker JR. Wheezing due to pulmonary embolism: treatment with heparin. N Engl J Med 1966;274: Windebank WJ, oyd G, Moran F. Pulmonary thromboembolism presenting as asthma. r Med J 1973;1: Lindsey HE, Wyllie JH. Release of prostaglandins from embolized lungs. r J Surg 1970;57: ustin JHM, Sagel SS. lterations of airway caliber after pulmonary embolization in the dog. Invest Radiol 1972;7: Remy-Jardin M, Remy J, Louvegny S, rtaud D, Deschildre F, Duhamel. irway changes in chronic pulmonary embolism: T findings in 33 patients. Radiology 1997;203: Garg K, Welsh H, Feyerabend J, et al. Pulmonary embolism: diagnosis with spiral T and ventilation perfusion scanning correlation with pulmonary angiographic results or clinical outcome. Radiology 1998;208: ergin J, Sirlin, Hauschildt JP, et al. hronic thromboembolism: diagnosis with helical T and MR imaging with angiographic and surgical correlation. Radiology 1997;204: Remy-Jardin M, Remy J, Wattinne L, Giraud F. entral pulmonary thromboembolism: diagnosis with spiral volumetric T with the single-breathhold technique comparison with pulmonary angiography. Radiology 1992;185: hartrand-lefebvre, Howarth N, Lucidarme O, et al. ontrast-enhanced helical T for pulmonary embolism detection: inter- and intraobserver agreement among radiologists with variable experience. JR 1999;172: JR:178, May