| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Report |
1
Department of Radiology, Vancouver Hospital and Health Sciences Centre,
University of British Columbia, 855 W. 12th Ave., Vancouver, B.C., V5Z 1M9,
Canada.
2
Department of Radiology, Osaka University Medical School, 2-2 Yamadaoka,
Suita, Osaka 565-0825, Japan.
Received February 26, 1999;
accepted after revision July 27, 1999.
Address correspondence to N. L. Müller.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CONCLUSION. High-resolution CT findings of pulmonary leukemic infiltrates reflect the predilection of leukemic cells to involve the perilymphatic pulmonary interstitium.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
All patients underwent high-resolution (1- to 3- mm collimation) CT at 10-mm intervals through the chest. In eight patients, images were prospectively targeted to either the right or left lung using a field of view of 19-25 cm. Images were reconstructed using a high-spatial-frequency algorithm and were photographed at lung windows (window width, 1000-1500 H; level from -550 to -700 H) and mediastinal windows (window width, 300-450 H; level, 10-35 H). One patient received IV contrast material for conventional CT performed before high-resolution CT.
CT scans were independently reviewed by two radiologists. Final decisions on the findings were determined by consensus. Radiologists paid particular attention to abnormalities located in biopsy regions; differences in findings for these locations compared with the remainder of the lungs were noted. For one patient, the site of transbronchial biopsy was unavailable. For all patients, the lungs were divided into three equal zones: upper, middle, and lower; and the zonal distribution of each finding was noted. Additionally, each finding was categorized as being central if it was located in the inner two thirds of the lung or peripheral if it was located in the outer one third.
Radiologists evaluated images for the presence of interlobular septal thickening, peribronchovascular interstitial thickening, nodules, ground-glass attenuation, and air-space consolidation. The interlobular septal thickening and the peribronchovascular thickening were classified as either smooth or nodular. Parenchymal nodules were classified as centrilobular, peribronchovascular, or random in distribution. Nodules were considered centrilobular if they were located in the region of the lobular core, adjacent to the centrilobular artery, or a few millimeters away from the interlobular septa or pleura. The size and number of the nodules were also assessed; nodules smaller than 1 cm were classified as small, and nodules larger than 1 cm were classified as large. Ground-glass attenuation was defined as parenchymal opacities that did not obscure adjacent vasculature; regions of consolidation were defined as parenchymal opacities that obscured underlying vessels.
Chest radiographs were reviewed retrospectively and independently of the high-resolution CT scans. Radiographs were available for five of the 10 patients; reports of the radiographs were available for three others. Neither the radiographs nor radiographic reports were available for two patients. Available radiographs were obtained at a median of 3 days (range, 1-30 days) before the high-resolution CT scans. The radiographs were evaluated for the presence of reticulation, consolidation, ground-glass opacities, nodules, and pleural effusion.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A notable finding on high-resolution CT was the thickening of the interlobular septa and bronchovascular bundles. Interlobular septal thickening was present in all patients (Fig. 1). The septal thickening was considered smooth in six patients, nodular in one, and smooth and nodular in three. In one patient, the only abnormality was smooth interlobular septal thickening and two small nodules. Thickening of the bronchovascular bundles was present in nine patients and was either smooth (n = 5) or nodular (n = 4) (Fig. 2A, 2B, 2C).
|
|
|
|
CT revealed parenchymal nodules within the lungs of all patients; five had fewer than 10 nodules each. Most nodules were small; the nodules in eight scans measured 5-10 mm in diameter, and the nodules in one scan measured less than 5 mm. Nodules were randomly distributed in three patients, were distributed primarily along the bronchovascular bundles in four, and had a centrilobular predominance in three (Figs. 2A, 2B, 2C and 3A, 3B). In one patient, confluence of poorly defined centrilobular nodules resulted in a focus of consolidation.
|
|
Areas of ground-glass attenuation and consolidation appeared in the scans of seven patients. In six patients, the regions of ground-glass attenuation were associated with areas of consolidation. In five patients, the consolidation was located in a predominantly peribronchial distribution (Fig. 4A, 4B) and was associated with air bronchograms. The opacities had a slight peripheral predominance in four patients (Fig. 5A, 5B) and a central predominance in three. Additionally, all scans containing consolidation showed either interlobular septal or peribronchovascular interstitial thickening.
|
|
|
|
The pattern of parenchymal abnormalities was influenced by the type of leukemia. Although two patients with acute myelogenous or lymphocytic leukemia showed smooth interlobular septal and peribronchovascular thickening, four patients with chronic myelogenous or lymphocytic leukemia showed nodular interstitial thickening. Smooth interstitial thickening was common in all patients with adult T-cell leukemia, but in none of these patients was interstitial thickening the dominant abnormality. In three patients, the dominant abnormality was focal areas of peribronchial consolidation, and in a fourth patient the predominant abnormality was poorly defined centrilobular nodules.
The interobserver agreement for the interpretation of the CT findings was 80% in the diagnosis of interlobular septal thickening, 80% in the diagnosis of nodules, 70% in the diagnosis of ground-glass attenuation, and 90% in the diagnosis of consolidation.
Radiologists reviewed the radiographs (n = 5) or radiology reports (n = 3) for eight patients and found parenchymal abnormalties on all images. Abnormalties consisted of fine reticular patterns (n = 4), multifocal areas of consolidation (n = 3), or ground-glass opacities (n = 1). Additional findings included septal Kerley B lines (n = 2), peribronchial thickening (n = 2), poorly defined nodules (n = 2), and pleural effusion (n = 1).
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Few studies describe the CT findings of pulmonary leukemic infiltration. Single case reports using conventional CT have described solitary [4] and multiple [5] pulmonary nodules resulting from leukemic infiltration. The CT findings of leukemic pleural plaques with associated pleural effusions have also been described [7].
We evaluated the findings of pulmonary leukemic infiltration on high-resolution CT. The most striking abnormality was interstitial thickening that involved both the axial and peripheral interstitium. This finding is consistent with pathology studies that describe a propensity for leukemic cells to infiltrate the peribronchial and peribronchiolar connective tissue; the cells then permeate the airway walls [8]. The peribronchial distribution of leukemic infiltration is probably caused by the predilection of leukemic cells for lymphatic routes [9].
Nodules were present in all patients; however, because of their small size and number, the nodules were not the dominant finding. The peribronchovascular distribution of nodules in four patients and the centrilobular nodules seen in three patients are probably related to the focal accumulation of leukemic cells in a perilymphatic distribution. Infiltration of leukemic cells in the adjacent air spaces probably accounted for the focal areas of consolidation with a peribronchial distribution [10, 11].
Abnormal patterns on chest radiographs varied and included reticulation, multifocal consolidation, ground-glass opacities, and small nodules. Half the radiographs showed a predominantly interstitial pattern, and two showed septal thickening in the form of Kerley B lines. Therefore, high-resolution CT scans were superior to radiographs in revealing leukemic infiltrates in a perilymphatic distribution.
Our study had several limitations. We studied patients with different forms of leukemia, including patients with adult T-cell leukemia, a form of leukemia that is relatively rare in North America. For nine patients, the diagnosis of pulmonary leukemic infiltration was made using a small sample of the lung parenchyma. Transbronchial lung biopsies, bronchoalveolar lavage, and open lung biopsies were directed at the sites of severe disease as revealed on CT; however, the abnormal patterns on CT at biopsy sites were similar to those found elsewhere in the lungs. It is conceivable, however, that other superimposed disease processes were present in the portions of the lungs that were not sampled. Therefore, some high-resolution CT findings might have been caused by other superimposed processes. We were able to obtain only limited information about abnormalities on radiographs and other information provided by high-resolution CT. However, based on available radiographs and radiology reports it seems reasonable to conclude that high-resolution CT scans were superior to radiographs in revealing the pattern and distribution of parenchymal abnormalities.
High-resolution CT findings of leukemic infiltrates are nonspecific. All findings have differential diagnoses; drug toxicity, pulmonary edema, hemorrhage, and infection can all have similar appearances.
In summary, pulmonary leukemic infiltration has a variety of findings on thin-section CT; however, we noted a distinct tendency for abnormalities to involve the perilymphatic interstitium. Although acute processes such as pulmonary edema and infection must be excluded, the presence of pulmonary leukemic infiltrates should be considered when high-resolution CT reveals interstitial thickening in patients with leukemia.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  D. J. Valentino III, M. C. Gelfand, V. E. Nava, D. F. Garvin, and A. H. Roberts II Hypoxemic Respiratory Failure in a 57-Year-Old Woman With Acute Monocytic Leukemia Chest, November 1, 2005; 128(5): 3629 - 3633. [Full Text] [PDF]
|
|
|
|
 |
|
 J. I. Jung, J. E. Choi, S. T. Hahn, C. K. Min, C. C. Kim, and S. H. Park Radiologic Features of All-Trans-Retinoic Acid Syndrome Am. J. Roentgenol., February 1, 2002; 178(2): 475 - 480. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
