DOI:10.2214/AJR.04.1847
AJR 2006; 186:368-373
© American Roentgen Ray Society
Drug-Induced Eosinophilic Pneumonia: High-Resolution CT Findings in 14 Patients
Carolina A. Souza1,
Nestor L. Müller1,
Takeshi Johkoh2 and
Masanori Akira3
1 Department of Radiology, Vancouver General Hospital, University of British
Columbia, 899 W 12th Ave., Vancouver V5Z 1M9, Canada.
2 Department of Radiology and Medical Physics, Osaka University Graduate School
of Medicine, Suita, Osaka, 565-0825, Japan.
3 National Kinki Chuo Hospital for Chest Disease, Sakai City, Osaka, 591-8555,
Japan.
Received December 3, 2004;
accepted after revision January 27, 2005.
Address correspondence to N. L. Müller.
Abstract
OBJECTIVE. The purpose of this study was to evaluate the
high-resolution CT findings of drug-induced eosinophilic pneumonias.
CONCLUSION. Drug-induced eosinophilic pneumonias usually manifest as
areas of consolidation and ground-glass opacity most commonly involving the
outer third of the lungs.
Keywords: chest • drug reaction • eosinophilic pneumonia • high-resolution CT • lung • lung diseases
Introduction
Eosinophilic lung diseases are a group of disorders characterized
pathologically by the presence of eosinophils in the alveoli and pulmonary
interstitium. They may be divided into those of unknown cause and those
related to a defined cause including parasitic infestation, fungal infection,
and drug reaction [1,
2].
Eosinophilic pneumonia is a well-recognized pattern of drug reaction.
Diagnosis is based on the presence of parenchymal opacification on the chest
radiograph, peripheral eosinophilia, and eosinophilic infiltration of the lung
parenchyma, associated with administration of a known drug and in the absence
of other causes of pulmonary disease
[3-8].
Early diagnosis is important because discontinuation of the offending drug
reduces morbidity and mortality.
High-resolution CT has an important role in the diagnosis of diffuse lung
diseases, allowing earlier detection and better characterization of the
abnormalities than does chest radiography
[1,
6,
7]. The high-resolution CT
manifestations of eosinophilic lung diseases have been described in several
studies [1,
2,
9]. However, the literature has
little information about the high-resolution CT findings of drug-related
eosinophilic pneumonias.
The purpose of the present study was to review the high-resolution CT
findings of biopsy-proven drug-induced eosinophilic pneumonia.
Materials and Methods
The study included 14 consecutive patients (nine men and five women; age
range, 29-81 years; mean, 53.3 years) with proven drug-induced eosinophilic
pneumonia who had undergone high-resolution CT of the chest at one of our
three institutions. The underlying disease and the offending drug for each
patient are summarized in Table
1.
View this table:
[in this window]
[in a new window]
|
TABLE 1: Causative Drug, Underlying Disease, Blood Eosinophil Count, and
High-Resolution CT (HRCT) Findings for Each Patient
|
|
All patients presented with nonspecific respiratory symptoms and pulmonary
opacities on radiographs during administration of a single drug (n
=12) or multiple drugs (n = 2). The two patients receiving
multiple-drug therapy were a patient with Hodgkin's disease and a patient with
tuberculosis. In these two patients, the drug responsible for the reaction was
identified by discontinuing the drug suspected of causing the eosinophilic
reaction and observing rapid patient improvement. The interval between the
initiation of drug therapy and the onset of symptoms ranged from 1 to 42 days
(mean, 8.7 days; median, 4 days) (Table
1). Diagnosis of drug-induced eosinophilic pneumonia was based on
clinical history and histologic findings obtained by transbronchial biopsy
(n = 10) or surgical biopsy (n = 4). Other infectious and
noninfectious causes of eosinophilic lung disease were excluded by laboratory
tests and clinical history. Blood eosinophilia (eosinophils > 5%) was
present in 12 (86%) of 14 patients.
High-resolution CT scans were obtained on a variety of scanners, using 1-
to 2-mm collimation at 10-mm intervals. The images were obtained with the
patient in the supine position at the end of inspiration and reconstructed
using a high-spatial-frequency algorithm. The high-resolution CT images
(window level, -600 to -700 H; window width; 1,000-1,500 H) were
retrospectively evaluated by two radiologists for the presence, extent, and
anatomic distribution of parenchymal abnormalities. The observers recorded the
predominant abnormality (ground-glass opacity, consolidation, or nodules) and
the associated findings (ground-glass opacity, consolidation, nodules,
interlobular septal thickening, or reticulation) in accord with the
definitions proposed by the Nomenclature Committee of the Fleischner Society
[10]. Nodules were classified
according to their size (< 5, 5-10, or > 10 mm), number (< 10 or
3 10), and anatomic distribution (centrilobular,
peribronchovascular, or random). The anatomic distribution of abnormalities
was classified as central if predominant in the inner third of the lungs,
peripheral if predominant in the outer third of the lungs, or random if there
was no predominance. Zonal predominance was classified as upper when most of
the abnormalities were above the level of the tracheal carina, lower when most
of the abnormalities were below this level, or random when no zonal
predominance was noted. The overall extent of abnormalities was classified as
involving less than 25%, 25-50%, 50-75%, or more than 75% of lung parenchyma.
Decisions about presence, distribution, and extent of abnormalities were
reached by consensus between the two radiologists. This retrospective study
was approved by the institutional review board of one of the three centers
participating in the study; no approval is required for retrospective studies
in the other two institutions.
The high-resolution CT examinations were performed within 2-30 days (mean,
11 days; median, 7 days) after the initiation of drug administration and
within 1-10 days (mean, 4 days; median, 3 days) after the onset of symptoms.
The interval between the high-resolution CT scans and the biopsy ranged from 1
to 7 days (mean, 3 days; median, 2 days).
Results
Table 2 summarizes the
high-resolution CT findings. All 14 patients had parenchymal abnormalities on
high-resolution CT scans, including air-space consolidation (n = 14)
and ground-glass opacities (n = 12). Air-space consolidation was the
predominant abnormality in 11 patients (79%) and ground-glass opacity in three
patients (21%) (Figs. 1 and
2).
View larger version (67K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2 —High-resolution CT scan of 81-year-old man receiving
5-aminosalicylic acid for colitis shows extensive bilateral air-space
consolidation and ground-glass opacity. Mild septal thickening is seen in
anterior aspects of both lungs (arrows).
|
|
Associated parenchymal findings included nodules, interlobular septal
thickening, and reticulation. Multiple nodules (> 10) were present in eight
(57%) of 14 patients, and all these nodules were of small size (< 5 mm) and
centrilobular distribution (Fig.
3A). Interlobular septal thickening, which was found in six (43%)
of 14 patients, was generally mild with no zonal predominance
(Fig. 3B). Bilateral
reticulation was seen in three patients (21%), in all of whom it was mild and
randomly distributed (Fig.
4A).
View larger version (133K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3A —31-year-old woman receiving minocycline for sinusitis.
High-resolution CT scan shows patchy ground-glass opacity and mild
consolidation in a peripheral distribution. Also seen are small centrilobular
nodules (straight arrows) and nodular thickening of lobular septa and
major fissure (curved arrows).
|
|
View larger version (101K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4A —50-year-old man receiving methotrexate therapy for non-Hodgkin's
lymphoma. High-resolution CT scan shows area of ground-glass opacity and
reticulation in periphery of left upper lobe. Also seen are a few small
nodules (arrows).
|
|
The abnormalities were bilateral and asymmetric in all patients and were
mainly in the peripheral lung regions in 10 patients (71%)
(Fig. 4B), in the central lung
regions in three (21%), and random in one (7%). The abnormalities involved
mainly the upper lung zones in seven patients (50%), the lower lung zones in
one (7%), and all lung zones to similar extent in six (43%). Five patients had
areas of air-space consolidation in a perilobular distribution surrounding an
area of ground-glass attenuation (Fig.
5). The overall extent of abnormalities was less than 25% of the
lung parenchyma in five patients, 25-50% in two, 50-75% in three, and more
than 75% in four.
View larger version (76K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4B —50-year-old man receiving methotrexate therapy for non-Hodgkin's
lymphoma. High-resolution CT scan at a lower level shows patchy areas of
consolidation and ground-glass opacity in a peripheral distribution.
|
|
View larger version (87K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5 —High-resolution CT scan of 47-year-old man receiving phenytoin for
seizures shows patchy bilateral air-space consolidation in a peripheral
distribution. Focal area of ground-glass opacity surrounded by crescent-shaped
consolidation ("reverse halo" sign) is seen in left lung
(arrow).
|
|
All 14 patients had a good outcome. In three patients the parenchymal
abnormalities cleared with discontinuation of the drug. In the remaining 11
patients, improvement was observed after corticosteroid therapy.
Discussion
Several studies have emphasized the importance of drug reactions
[3-8,
11,
12]. The lung is one of the
most commonly involved organs, and drug-induced pulmonary injury may be an
important cause of morbidity and mortality. Drug reactions in the lung may
result in a variety of histologic patterns, including diffuse alveolar damage,
nonspecific interstitial pneumonia, bronchiolitis obliterans with organizing
pneumonia, pulmonary hemorrhage, and eosinophilic pneumonia
[3-8].
Classification of drug-induced lung diseases according to the pathologic
pattern is useful because clinical and radiologic manifestations often reflect
the underlying histology [7,
13].
Eosinophilic pneumonia is a common pattern of drug-related lung injury
[3-8,
11-13]
and may be caused by a number of agents. As of November 2004, approximately
115 drugs had been reported as causes of eosinophilic lung disease
[14]. The most common are
summarized in Table 3. In our
series, nine of 14 patients were receiving drugs with a relatively commonly
associated eosinophilic reaction and four patients were receiving drugs
described as causing an eosinophilic drug reaction in only a few cases
(ethambutol, sodium cromoglycate, cefaclor, and 5-aminosalicylic acid)
[7]. We describe a case of
eosinophilic reaction secondary to aldose reductase inhibitor, an association
to our knowledge not previously reported
(Table 1).
Drug-induced eosinophilic pneumonias may be acute in onset or may have an
insidious course with progression for months after therapy institution
[4,
8]. Lung injury is unrelated to
the cumulative dose. Acute disease is usually mild and self-limited. In the
chronic form, presentation varies from minimally symptomatic to severe
illness, with high fever, dyspnea, chills, malaise, and weight loss
[4,
15]. Prognosis is good if the
offending drug is stopped; in some cases, however, treatment with
corticosteroids is required [8,
13]. In our series, three
patients improved with discontinuation of the causative drug and 11 required
corticosteroid therapy. Both acute and chronic forms are characterized
pathologically by infiltration of the alveolar air spaces and interstitium
with eosinophils and lymphocytes
[3,
4,
7,
12]. Biopsy is often required
to establish the definitive diagnosis. In clinical practice, however,
diagnosis is usually based on clinical symptoms, the presence of parenchymal
opacities on the chest radiograph or CT scan, and increased eosinophils in the
blood or in the bronchoalveolar lavage fluid. Lymphocyte stimulation testing
is frequently positive and may be helpful in the diagnosis
[1,
2,
16]. Although characteristic,
the presence of peripheral eosinophilia is quite variable, having been
described in approximately 40% of patients
[13,
17,
18]. In our series, 86% of
patients had blood eosinophilia.
Air-space consolidation, areas of ground-glass attenuation, nodules, and
irregular lines have been described as high-resolution CT manifestations of
drug-induced eosinophilic pneumonias
[1,
7,
9,
19]. In our series, bilateral
areas of consolidation and ground-glass opacities were the most common
manifestations, present in all patients. A predominantly peripheral
distribution was seen in 72% of patients, involving mainly the upper lung
zones in 50%. Ancillary findings included small centrilobular nodules, present
in 57% of patients; interlobular septal thickening; and reticulation.
In comparison with the previously published data on the CT appearance of
eosinophilic lung diseases, the majority of our patients (10/14) showed
findings resembling those of chronic eosinophilic pneumonia— that is,
bilateral air-space consolidation and ground-glass opacities in a peripheral
distribution, often involving the middle and upper lung zones
[1,
7,
20-22].
Poorly defined centrilobular nodules and septal lines have been described more
commonly in acute eosinophilic pneumonitis, in which parenchymal opacities
tend to be extensive and do not have a zonal predominance
[1,
2,
9,
19,
23]. The remaining patients
(4/14) in the current study showed predominantly central or random
distribution of ground-glass opacities or consolidation.
In five patients, areas of air-space consolidation had a perilobular
distribution appearing as a crescent-shaped dense consolidation surrounding an
area of ground-glass attenuation. This finding resembles the "reverse
halo" sign described in association with bronchiolitis obliterans with
organizing pneumonia [24]. A
reaction similar to bronchiolitis obliterans with organizing pneumonia may be
associated with a variety of pathologic processes, including drug reactions
[12]. In the current study,
however, histologic findings showed only eosinophilic pneumonia.
Our study had several limitations. It was retrospective and included few
patients. Although diagnosis was proven by biopsy, in the majority of patients
this was limited to transbronchial biopsy. Therefore, although the biopsy
specimens showed findings consistent only with an eosinophilic drug reaction,
it is possible that other types of reaction may have been missed.
In summary, the high-resolution CT findings of drug-induced eosinophilic
pneumonias, consisting of air-space consolidation and ground-glass attenuation
in a predominantly peripheral distribution, are similar to those of
eosinophilic pneumonias of other causes. The diagnosis of drug-induced injury
requires a high index of suspicion and careful clinical correlation, a history
of exposure to a drug known to induce eosinophilia, and exclusion of other
causes of eosinophilic lung disease.
References
- Johkoh T, Müller NL, Akira M, et al. Eosinophilic lung
diseases: diagnostic accuracy of thin-section CT in 111 patients.
Radiology 2000;216
: 773-780[Abstract/Free Full Text]
- Allen JN, Davis WB. Eosinophilic lung diseases. Am J
Respir Crit Care Med 1994;150
: 1423-1438[Medline]
- Pietra GG. Pathologic mechanisms of drug-induced lung disorders.
J Thorac Imaging 1991;6
: 1-7[Medline]
- Erasmus JJ, McAdams HP, Rossi SE. Drug-induced lung injury.
Semin Roentgenol 2002;37
: 72-81[CrossRef][Medline]
- Smith GJ. The histopathology of pulmonary reactions to drugs.
Clin Chest Med 1990;11
: 95-117[Medline]
- Kuhlman JE. The role of chest computed tomography in the diagnosis
of drug-related reactions. J Thorac Imaging1991; 6:52
-61[Medline]
- Cleverley JR, Screaton NJ, Hiorns MP, Flint JD, Müller NL.
Drug-induced lung disease: high-resolution CT and histological findings.
Clin Radiol 2002;57
: 292-299[CrossRef][Medline]
- Rossi SE, Erasmus JJ, McAdams HP, Sporn TA, Goodman PC. Pulmonary
drug toxicity: radiologic and pathologic manifestations.
RadioGraphics 2000;20
: 1245-1259[Abstract/Free Full Text]
- Kim Y, Lee KS, Choi DC, Primack SL, Im JG. The spectrum of
eosinophilic lung disease: radiologic findings. J Comput Assist
Tomogr 1997; 21:920
-930[CrossRef][Medline]
- Austin JH, Müller NL, Friedman PJ, et al. Glossary of terms
for CT of the lungs: recommendations of the Nomenclature Committee of the
Fleischner Society. Radiology 1996;200
: 327-331[Free Full Text]
- Erasmus JJ, McAdams HP, Rossi SE. High-resolution CT of
drug-induced lung disease. Radiol Clin North Am2002; 40:61
-72[CrossRef][Medline]
- Rosenow EC, Myers JL, Swensen SJ, Pisani RJ. Drug-induced pulmonary
disease: an update. Chest 1992;102
: 239-250[Free Full Text]
- Müller NL, White DA, Jiang H, Gemma A. Diagnosis and
management of drug-associated interstitial lung disease. Br J
Cancer 2004; 91[suppl 2]:S24
-S30
- Foucher P, Camus P. Pneumotox on line: the drug-induced lung
diseases. Available at:
www.pneumotox.com.
Accessed July 31, 2005
- Carrington CB, Addington WW, Goff AM, et al. Chronic eosinophilic
pneumonia. N Engl J Med 1969;280
: 787-798[Medline]
- Allen JN, Pacht ER, Gadek JE, Davis WB. Acute eosinophilic
pneumonia as a reversible cause of noninfectious respiratory failure.
N Engl J Med 1989;321
: 569-574[Abstract]
- Cooper JA Jr, White DA, Matthay RA. Drug-induced pulmonary disease.
Part 2. Noncytotoxic drugs. Am Rev Respir Dis1986; 133:488
-503[Medline]
- Cooper JAD, White DA, Matthay RA. Drug-induced pulmonary disease.
Part 1. Cytotoxic drugs. Am Rev Respir Dis1986; 133:321
-340[Medline]
- Padley SP, Adler B, Hansell DM, Müller NL. High-resolution
computed tomography of drug-induced lung disease. Clin
Radiol 1992; 46:232
-236[CrossRef][Medline]
- Mayo JR, Müller NL, Road J, Sisler J, Lillington G. Chronic
eosinophilic pneumonia: CT findings in six cases. AJR1989; 153:727
-730[Abstract/Free Full Text]
- Ebara H, Ikezoe J, Johkoh T, et al. Chronic eosinophilic pneumonia:
evolution of chest radiograms and CT features. J Comput Assist
Tomogr 1994; 18:737
-744[Medline]
- Arakawa H, Kurihara Y, Niimi H, Nakajima Y, Johkoh T, Nakamura H.
Bronchiolitis obliterans with organizing pneumonia versus chronic eosinophilic
pneumonia: high-resolution CT findings in 81 patients.
AJR 2001; 176:1053
-1058[Abstract/Free Full Text]
- Tomiyama N, Müller NL, Johkoh T, et al. Acute parenchymal lung
disease in immunocompetent patients: diagnostic accuracy of high-resolution
CT. AJR 2000; 174:1745
-1750[Abstract/Free Full Text]
- Kim SJ, Lee KS, Ryu YH, et al. Reversed halo sign on
high-resolution CT of cryptogenic organizing pneumonia: diagnostic
implications. AJR 2003;180
: 1251-1254[Abstract/Free Full Text]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
|
|
 
Y. J. Jeong, K.-I. Kim, I. J. Seo, C. H. Lee, K. N. Lee, K. N. Kim, J. S. Kim, and W. J. Kwon
Eosinophilic Lung Diseases: A Clinical, Radiologic, and Pathologic Overview
RadioGraphics,
May 1, 2007;
27(3):
617 - 637.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
