| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, Hospital de Sant Pau, Universitat Autónoma de
Barcelona, Avda Sant Antoni Ma Claret 167, 08025 Barcelona,
Spain.
2 Department of Hematology, Hospital de Sant Pau, Universitat Autónoma de
Barcelona, 08025 Barcelona, Spain.
Received March 14, 2005;
accepted after revision July 27, 2005.
Address correspondence to T. Franquet.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. The study population consisted of 26 recipients of hematopoietic stem cell transplants who had proven respiratory viral pneumonia. In all cases, thin-section CT scans were obtained before fiberoptic bronchoscopy and bronchoalveolar lavage. The study included only patients in whom bronchoalveolar lavage fluid showed no evidence of organisms other than respiratory viruses. The CT scans were assessed for the presence, extent, and anatomic distribution of ground-glass attenuation, air-space consolidation, nodules, centrilobular branching structures (tree-in-bud), thickening of the bronchovascular bundles, and pleural effusion.
RESULTS. Areas of ground-glass attenuation were identified in 24 (92%) of 26 patients and were the only finding in eight patients. Multiple nodules, seen in 17 (65%) of 26 patients, measured 3-10 mm in diameter or were larger than 10 mm. The nodules had a centrilobular or random distribution. A tree-in-bud appearance was seen in six of the patients with centrilobular nodules. This pattern had a bilateral distribution and involved mainly the lower lung zones. CT revealed thickening of the bronchovascular bundles in 16 (61%) of the patients. Thickening was bilateral in 14 and unilateral in two patients. Bronchial wall thickening involved the lower lobes in six patients and had a patchy random distribution in the remaining nine patients. Air-space consolidation was present in nine (35%) of the cases. It had a lobular or subsegmental distribution in eight of the patients and a segmental distribution in one patient. Areas of consolidation were randomly distributed throughout the lungs in all cases. Less common findings included bilateral pleural effusion and bronchial dilatation.
CONCLUSION. Respiratory viral infection is common among adult recipients of hematopoietic stem cell transplants, occurring over a wide time span after transplantation. The presence of respiratory viral infection must be considered in any patient with new respiratory symptoms, fever, or findings at CT such as extensive or patchy areas of ground-glass opacities or a mixture of patterns, most commonly ground-glass attenuation, thickening of the bronchial walls, and multiple small nodules.
Keywords: CT • infectious disease • lung • transplantation • viral pneumonia
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Radiologic descriptions of the manifestations of respiratory viral infections in HSCT recipients have been limited to studies including only small numbers of patients. Thin-section CT of the chest is considered the primary imaging technique in the early diagnosis of opportunistic infections in HSCT recipients [2-4]. The aim of this study was to assess the incidence of respiratory viral infection and lung abnormalities on serial thin-section CT scans in a large patient population at high risk of pulmonary complications.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Thin-Section CT Scans
In all cases, thin-section CT scans were obtained before fiberoptic
bronchoscopy and BAL. CT scans were obtained with a Somaton Plus 4 CT scanner
(Siemens Medical Solutions), a Toshiba 900 CT scanner, or a Toshiba Asteion CT
scanner (Toshiba Medical Systems). The scans were obtained at end inspiration
with 1.0- or 2-mm collimation and obtained at 10- or 20-mm intervals from the
apex of the lung to the diaphragm. The scans were obtained with 120 kV and
200-320 mA. All images were interpreted and photographed at window settings
appropriate for pulmonary parenchyma (level, -700 H; width, 900 H) and
mediastinum (level, 30-50 H; width, 350-500 H). The thin-section CT findings
at presentation in five patients in the current study were previously reported
[7].
Image Interpretation
Hard copies of the thin-section CT images were jointly reviewed by two
experienced thoracic radiologists who were aware that all patients had proven
pulmonary viral infection; however, the radiologists were unaware of other
clinical information. A conclusion was reached by consensus if there was
disagreement.
The CT scans were assessed for the presence, extent, and anatomic distribution of ground-glass attenuation, air-space consolidation, nodules, centrilobular branching structures (tree-in-bud), thickening of the bronchovascular bundles, and pleural effusion. Ground-glass attenuation was defined as an area of hazy increased attenuation without obscuration of underlying vascular markings. Air-space consolidation was considered present when the opacities obscured the underlying vessels. Parenchymal nodules were subcategorized according to their distribution (centrilobular, peribronchovascular, subpleural, random), location (upper or lower lung zone), marginal characteristics (smooth or irregular), and number (single or multiple). Centrilobular nodules were defined as parenchymal opacities located in the central portion of the secondary pulmonary lobule (lobular core). Among the patients with a centrilobular distribution, the presence or absence of a tree-in-bud appearance— that is, branching linear and nodular opacities resembling a branching tree—also was noted. Nodules also were assessed as being surrounded by a halo of ground-glass attenuation. The presence of associated findings, such as pleural effusion, also was assessed.
Anatomic distribution was regarded as predominantly peripheral if abnormalities were seen mostly in the outer third of the lung, central if most were in the inner third of the lung, peribronchial if abnormalities occurred along the bronchovascular bundles, and random otherwise. The distribution was further classified as lobar, segmental, or patchy or nonsegmental. Zonal predominance was divided into upper and lower. Upper lung zone predominance was considered present when most of the abnormalities were above the level of the tracheal carina; lower zone predominance was considered present when the abnormalities were below that level.
Patients with pulmonary edema, hemorrhage, and recurrent or ongoing malignancy were excluded from the study. If the ground-glass pattern was present, other causes, such as aspiration, graft-versus-host disease, hyperhydration, or other infections, were ruled out on the basis of clinical factors (type of bone marrow transplantation, increasing weight, central venous pressure measurements), microbiologic data, or a combination of these factors.
Microbiologic Sampling and Analysis
Thin-section CT, fiberoptic bronchoscopy, and BAL were performed within 24
hours of one another. BAL procedures were performed according to a
standardized protocol. Accordingly, five boluses of 30-50 mL of sterile saline
solution were instilled into the distal bronchial tree at the site of the CT
abnormality. The last four instillations were used for cytologic and
microbiologic analysis, whereas the first one was discarded. Gram, acridine
orange, auramine, and Giemsa stains were used for direct identification of
bacteria, mycobacteria, fungi, and parasites, respectively. Polymerase chain
reaction (PCR) assays were performed for detection of Mycoplasma
pneumoniae, Chlamydophila pneumoniae, and Legionella
pneumophila. All bronchoscopic procedures were performed at the bedside
with local anesthesia and supplemental oxygen.
In all cases, diagnosis of infection required a positive viral culture result or direct antigen detection by DFA test, reverse transcriptase PCR (RT-PCR), or both performed on BAL fluid. The study included only patients in whom BAL fluid showed no evidence of organisms other than respiratory viruses. BAL specimens were processed within 2 hours of sampling for the presence of respiratory viruses through antigen detection by immunofluorescence study, viral culture, RT-PCR, or a combination of these procedures. Qualitative RT-PCR assays were performed on all specimens for the following respiratory viruses: influenza A and B, respiratory syncytial virus (RSV), parainfluenza virus (PIV) 1 and 3, and human metapneumovirus.
Clinical records were reviewed by two of the authors for signs and symptoms, predisposing conditions, and clinical course. The diseases for which HSCT was performed were acute myeloid leukemia (n = 9), Hodgkin's disease (n = 4), myelodysplastic syndrome (n = 4), non-Hodgkin's lymphoma (n = 3), multiple myeloma (n = 3), acute lymphocytic leukemia (n = 2), and chronic lymphatic leukemia (n = 1). This study was conducted with institutional review board approval; patient informed consent was not required.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although respiratory viral infection is seen with increased frequency in HSCT recipients, the prevalence varies considerably among studies [4, 8]. In a prospective study conducted by the European Group for Blood and Marrow Transplantation, 40 respiratory viral infections (2%) were diagnosed in 1,863 patients [5]. Whimbey and Ghosh [11] evaluated the role of respiratory viral infections in HSCT recipients and found an incidence of RSV pneumonia of 9.2%. In another study, Leung et al. [4] reviewed pulmonary infections in 59 bone marrow transplant recipients. Although Cytomegalovirus (n = 22 [37%]) and Aspergillus (n = 17 [29%]) were the two most common pathogens, respiratory viral infections were uncommon and occurred in only three (5%) of the 59 infectious episodes (two of RSV and one of influenza B infection).
Unlike previous investigators, we found a high prevalence of respiratory viral infection (20%) among our patients. The difference between our prevalence and that previously reported was probably due to underestimation of the true incidence rate, because direct antigen detection with DFA test or RT-PCR analysis performed on BAL fluid was not previously systematically used in evaluation of all HSCT recipients with symptoms [12]. Therefore respiratory viruses may be responsible for a large number of cases of pneumonia previously classified as idiopathic [10].
Influenza A virus accounted for most of the respiratory viral infections in our series (n = 11); the other cases were caused by metapneumovirus (n = 5), PIV 3 (n = 4), RSV (n = 3), and adenovirus (n = 3). These respiratory viruses are the most commonly found viruses in studies involving HSCT recipients [5, 6, 9]. However, the relative proportions vary among studies, probably because of the epidemic in the population during the study period. This specific seasonal characteristic is also illustrated by a relatively high frequency of influenza A infection in our series. A large epidemic of influenza A virus infection occurred during the study period, and the isolates recovered from patients closely reflected the respiratory viruses isolated in the community. In our series, nosocomial respiratory viral infection was identified in five (19%) of 26 patients.
Most respiratory viral infections produce acute symptoms such as fever, nonproductive cough, dyspnea, and hypoxemia. Rapid detection methods are essential when HSCT recipients have fever along with signs, symptoms, or radiographic indications of pulmonary infection. In our series, every patient had clinical signs or symptoms of acute infection. Respiratory viral infection was diagnosed with systematic use of antigen detection by DFA test, viral culture, or sensitive RT-PCR assays on appropriate samples such as BAL specimens [13].
Conventional chest radiography is the first imaging technique performed on HSCT recipients with fever. However, an increasing number of patients undergo thin-section CT when there is high clinical suspicion of pneumonia with normal or questionable radiographic findings. Several studies have shown that thin-section CT is the most sensitive imaging method for detecting early lung changes in immunocompromised patients with acute pulmonary diseases [3, 14-16]. In a study of 87 consecutively treated patients with febrile neutropenia, Heussel et al. [16] found that in 50% of subjects, CT revealed a pulmonary lesion not seen on radiographs.
The descriptions of the thin-section CT appearance of respiratory viral infections have been limited to very few studies. Oikonomou et al. [17] reviewed the thin-section CT findings in four patients with hematologic malignant disease and influenza A pneumonitis and found that the predominant CT findings were ground-glass opacities and centrilobular nodules smaller than 10 mm in diameter. Gasparetto et al. [18] reviewed the thin-section CT findings in 20 patients with RSV pneumonitis after HSCT and found that the most common thin-section CT findings were small centrilobular nodules and multifocal areas of consolidation and ground-glass opacities in a bilateral asymmetric distribution. Franquet et al. [7] reviewed the CT findings in five patients with human metapneumovirus pneumonia and found that the predominant CT findings were patchy areas of ground-glass attenuation, small nodules, and multifocal areas of consolidation in a bilateral asymmetric distribution.
In the present study, ground-glass attenuation was found in 24 (92%) of the patients. The areas of ground-glass attenuation were usually bilateral and had a patchy random distribution. Furthermore, ground-glass attenuation was the only observed abnormality in eight (31%) of the patients. Unfortunately, these findings are nonspecific and can be found in other types of bacterial or viral pneumonia, such as herpes simplex and cytomegalovirus infections [19-23]. Ground-glass attenuation is known to be a pattern of low specificity and can reflect disturbances of ventilation, temporary overhydration, aspiration, graft-versus-host disease, and disturbances of perfusion. All these clinical circumstances were carefully ruled out in all the episodes reported in this study. Moreover, the fact that high-resolution CT changes were bilateral makes the diagnosis of infectious pneumonia associated with ground-glass opacities more reliable. In agreement with previous reports on the high-resolution CT appearance of various viral infections, we showed that the bilateral and patchy distribution of ground-glass attenuation is the most common pattern of respiratory viral infection, regardless of cause.
In the current study, thickening of the bronchial walls was found in 16 patients. This finding was not surprising because viral airway infection leads to inflammatory changes along the bronchovascular bundles, resulting in airway wall thickening of the bronchi and bronchioles. A combination of ground-glass attenuation and bronchial wall thickening was seen in 11 patients.
A tree-in-bud pattern was found in six patients in our series. Typically, the thin-section CT findings consist of centrilobular opacities arranged in a tree-in-bud pattern, manifested by small Y- and V-shaped opacities in the lung periphery. These opacities represent bronchioles impacted by inflammatory secretions. Infectious disorders involving the small airways are the most common causes of this sign [24]. Although bacteria are the most common cause of infectious bronchiolitis in immunocompromised patients, the differential diagnosis of this pattern includes mycobacterial, viral, and fungal infections [2, 21].
Small nodular opacities have been found in patients with CMV, herpes simplex, and herpes varicella-zoster pulmonary infections [5, 6, 10, 14]. We found small nodules ranging in size from 3 to 10 mm in 17 (65%) of 26 patients. Other associated CT findings were the absence of cavitation (100%), the presence of ill-defined borders (92%), and the presence of a halo of ground-glass attenuation (52%). Although used to describe the CT appearance of hemorrhagic nodules due to invasive aspergillosis [25], the halo sign has been shown to occur in association with hemorrhagic nodules of varying causes [26]. The focal or multifocal areas of air-space consolidation seen in nine of our patients were indistinguishable from those obtained in patients with other forms of viral or bacterial bronchopneumonia.
The range of thin-section CT findings in our series resembled the spectrum of radiologic findings of respiratory viral infections reported by others [4, 17, 18]. These findings usually consist of a mixture of patterns, most commonly ground-glass attenuation, thickening of the bronchovascular bundles, multiple small nodules, and air-space consolidation.
Our study had several limitations. First, although it included a larger number of patients than previous studies, the study was a descriptive, retrospective review of findings. Second, because it was a retrospective study in which clinical and radiologic records were used to identify patients, it is possible that some patients with respiratory viral infections were not included. Last, the study findings did not allow conclusions about the accuracy of thin-section CT in differentiating respiratory viral infection from other diseases. That would require blind comparison of thin-section CT findings in respiratory viral infections with findings in other diseases.
Although thin-section CT findings in respiratory viral infection are nonspecific and the same agent can produce more than one pattern, radiologists should be aware of the more frequent findings associated with the more commonly encountered infections. Because we did not compare imaging of respiratory viral infections with that of other infections, we are unable to comment on the predictive value of our findings. In clinical practice, the differential diagnosis must include entities other than viral infections. Thus even though viral infection may be suggested, radiologists need to consider many noninfectious and infectious disorders that should be differentiated from respiratory viral infections in daily clinical practice. A combined strategy consisting of thin-section CT and guided bronchoscopy with BAL performed within the first 24 hours after CT may improve the diagnostic yield and subsequent therapeutic changes in these patients.
The main goal of the study was to describe the thin-section CT features of respiratory viral infections in a large group of HSCT recipients. The current series, which included 26 HSCT recipients infected with both community-acquired and nosocomial respiratory viruses is, to our knowledge, the largest series to date from a single institution. Our experience emphasizes that respiratory viruses, especially when such viruses are circulating in the community, can be the origin of significant infectious episodes in HSCT recipients.
In conclusion, respiratory viral infection is common among adult HSCT recipients, occurring over a wide time span after transplantation. Respiratory viral infection must be considered in any patient with new respiratory symptoms, fever, or certain findings at CT, such as extensive or patchy areas of ground-glass opacities, or a mixture of patterns, most commonly ground-glass attenuation, thickening of the bronchial walls, and multiple small nodules. When present, these findings should alert the clinician to the possibility of respiratory viral infection and prompt the referring physician to perform the other studies necessary to establish the diagnosis and treatment plan. A tandem strategy consisting of BAL guided by previous thin-section CT findings may improve the diagnostic yield and subsequent therapeutic changes in these patients.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
