Published online before print November 22, 2005, 10.1148/radiol.2381040088
(Radiology 2005;238:330-338.)
© RSNA, 2005
Chlamydia Pneumoniae: Comparison with Findings of Mycoplasma Pneumoniae and Streptococcus Pneumoniae at Thin-Section CT1
Atsushi Nambu, MD,
Akitoshi Saito, MD,
Tsutomu Araki, MD,
Katsura Ozawa, MD,
Yoshimitsu Hiejima, PhD2,
Masaki Akao, MD,
Zennosuke Ohki, MD and
Hiroshi Yamaguchi, MD
1 From the Departments of Radiology (A.N.) and Internal Medicine (K.O., M.A., Z.O., H.Y.), Kofu Municipal Hospital, Kofu, Yamanashi prefecture, Japan; Department of Radiology, University of Yamanashi (A.S., T.A.), Shimokawato 1110, Tamaho-cho, Nakakoma-gun, Yamanashi prefecture, Japan 400-3898; and Center for Life Science Research, Bioinformatics Support Section, Data Science Division, University of Yamanashi, Yamanashi, Japan (Y.H.). Received January 21, 2004; revision requested March 19; revision received February 5, 2005; accepted March 2.
Address correspondence to A.N. (e-mail: nambu-a{at}gray.plala.or.jp).
 |
ABSTRACT
|
|---|
Purpose: To retrospectively compare thin-section computed tomographic (CT) findings of Chlamydia pneumoniae pneumonia with those of Streptococcus pneumoniae pneumonia and Mycoplasma pneumoniae pneumonia.
Materials and Methods: Institutional review board and patient informed consent were not required. Twenty-four patients with C pneumoniae pneumonia (17 men, seven women; age range, 19–89 years) underwent thin-section CT; 41 patients with S pneumoniae pneumonia (28 men, 13 women; age range, 19–91 years) and 30 patients with M pneumoniae pneumonia (20 men, 10 women; age range, 16–67 years) were also enrolled. Thin-section CT scans of each patient were retrospectively and independently assessed by two chest radiologists for consolidation, ground-glass opacity (GGO), bronchovascular bundle thickening, nodules, pleural effusion, lymphadenopathy, reticular or linear opacity, airway dilatation, pulmonary emphysema, and bilateral lung involvement. Consensus was reached for disagreements. The frequency of each finding was compared among the three types of pneumonia by using the 
2 test.
Results: For C pneumoniae pneumonia, CT demonstrated consolidation in 20 patients, GGO in 13, bronchovascular bundle thickening in 17, nodules in 18, pleural effusion in six, lymphadenopathy in eight, reticular or linear opacity in 15, airway dilatation in nine, pulmonary emphysema in 11, and bilateral lung involvement in 12. Bronchovascular bundle thickening (P = .022) and airway dilatation (P = .034) were significantly more frequent in patients with C pneumoniae pneumonia than in those with S pneumoniae pneumonia. Reticular or linear opacity (P = .017), airway dilatation (P = .016), and associated pulmonary emphysema (P = .003) were significantly more frequent in patients with C pneumoniae pneumonia than in those with M pneumoniae pneumonia.
Conclusion: C pneumoniae pneumonia demonstrates a wide spectrum of thin-section CT findings that are similar to those of S pneumoniae pneumonia and M pneumoniae pneumonia; airway dilatation and bronchovascular thickening were significantly more frequent in patients with C pneumoniae pneumonia.
© RSNA, 2005
|
INTRODUCTION
|
|---|
Chlamydia pneumoniae was reported as a pathogen of human pneumonia in 1989 and was determined to be distinctive from other Chlamydia species, such as Chlamydia trachomatis or Chlamydia psittaci (1). Previous development of serologic tests for this microorganism has enabled researchers to diagnose C pneumoniae pneumonia and has revealed that 6%–12% of cases of community-acquired pneumonia are caused by C pneumoniae (1–3). C pneumoniae is now the third most common pathogen after Streptococcus pneumoniae and Mycoplasma pneumoniae in community-acquired pneumonia (4). Pneumonia that is cause by pathogens of the Chlamydia species, together with M pneumoniae, is known as atypical pneumonia, which is clinically characterized by a nonproductive cough or a mildly elevated or normal white blood cell count. Atypical pneumonia requires a different treatment strategy than usual bacterial pneumonia, such as streptococcal pneumonia, and therefore the correct diagnosis of atypical pneumonia is clinically important (3).
Although the clinical picture of atypical pneumonia is somewhat different than that of usual bacterial pneumonia, this difference is often not definitive for differential diagnosis (4). The serologic diagnosis usually requires several days and cannot be completed in time for decision making of initial medication, which is critical to the treatment of community-acquired pneumonia. Also, there are notable differences in the clinical manifestation of C pneumoniae pneumonia and M pneumoniae pneumonia. Such differences should allow for the distinction between these two types of pneumonia, despite similarities in effective antibiotics. First, M pneumoniae pneumonia occasionally results in complications of the central nervous system, such as meningitis or transverse myelitis (5); these complications are rare in patients with C pneumoniae pneumonia. Second, only C pneumoniae has been noted to be associated with coronary artery disease (3). Third, C pneumoniae often leads to reinfection or chronic infection in patients who carry C pneumoniae antibodies and occasionally requires a prolonged course of medication to eradicate the disease (3). These differences potentially affect the management of each disease.
Hence, radiologic examinations are expected to provide information about the extent and differential diagnosis of pneumonia.
A few reports have focused on the radiologic features of C pneumoniae pneumonia (2,6,7). To the best of our knowledge, however, no specific imaging features for C pneumoniae pneumonia have been found on chest radiographs to date. The purpose of our study was to retrospectively compare the thin-section computed tomographic (CT) findings of C pneumoniae pneumonia with those of S pneumoniae pneumonia and M pneumoniae pneumonia.
|
MATERIALS AND METHODS
|
|---|
Patients and Methods of Diagnosis
The institutional review board at Kofu Municipal Hospital does not require approval for a retrospective study of routinely obtained clinical data. Patient informed consent was not required.
Between May 1999 and October 2002, there were 397 patients with community-acquired pneumonia in our hospital. The criteria for community-acquired pneumonia included respiratory symptoms, such as cough or dyspnea with fever, and abnormalities on chest radiographs or CT scans that were compatible with pneumonia in outpatients or inpatients who had been hospitalized within 48 hours.
Routine clinical examinations for patients with community-acquired pneumonia consisted of sputum Gram staining and sputum culture for both acid fast and conventional bacteria. Antibody titers were determined for M pneumoniae, and C pneumoniae antibodies, including immunoglobulin A and immunoglobulin G, were measured by using an enzyme-linked radioimmunoassay kit (Elisa; Hitachi Chemical, Shinjuku-ku, Japan). Chest radiography was also performed.
In addition, chest CT was routinely performed. We found that the percentage of patients with noninfectious pneumonia, such as bronchiolitis obliterans with organizing pneumonia, was higher than expected and that nearly 10% of patients with respiratory symptoms in our institution (Hiroshi Yamaguchi, oral communication, November 2002) had noninfectious pneumonia. This was partly because our institution was a referral hospital, and therefore many patients who had refractory symptoms, even if such patients were taking oral antibiotics, were referred to our hospital by family doctors. Thin-section CT often provided useful additional information regarding the distinction between infectious and noninfectious pneumonia because thin-section CT findings of noninfectious pneumonia are well known. To avoid treatment delay and to justify bronchoscopy with bronchoalveolar lavage in patients with noninfectious pneumonia, we incorporated chest CT into the routine work-up of patients suspected of having pneumonia. Although clinical data and chest radiographic findings are often enough to discriminate between the two types of pneumonia, we believe that a higher diagnostic accuracy must be achieved to warrant invasive examination, such as bronchoscopy, or steroid administration. In patients with slight clinical symptoms, however, chest CT was omitted.
An arterial blood culture was occasionally performed in patients whose temperature exceeded 38.5°C. We obtained arterial blood rather than venous blood for blood culture because arterial blood culture may allow for a more sensitive detection of some bacteria. Urine Legionella antigen was measured in patients with a recent history of bathing in a hot spring, which is the major source of Legionella infestation in Japan. Several viral antibody tests were also occasionally performed during epidemics of certain viral infections or in patients with normal white blood cell counts; these tests, however, were not routine.
We used the diagnostic criteria for acute infection of C pneumoniae that was established by Kishimoto et al (8,9) by using the enzyme-linked radioimmunoassay kit; C pneumoniae pneumonia was considered to be present when the immunoglobulin A or immunoglobulin G index exceeded 3.00 or when there was an interval increase (ie, the elevation of a figure of the index during paired tests) of more than 1.00 in immunoglobulin A or 1.35 in immunoglobulin G for paired serum specimens. When compared with previous methods, these criteria proved to be more accurate for the detection of acute C pneumoniae and yielded a specificity of 93.4% (ie, 6.6% of the healthy population showed a cutoff value above this level) and a sensitivity of 64.9% (9). It was documented that cross-reaction with other Chlamydia species was less frequent with these criteria than with the previous method, with cross-reaction occurring in one of four cases of Chlamydia psittaci and in one of 32 cases of Chlamydia trachomatis (8).
M pneumoniae pneumonia was diagnosed when the antibody titer was more than 160-fold greater than normal. S pneumoniae pneumonia was diagnosed if the sputum or blood culture was positive.
Of 397 patients with community-acquired pneumonia, 48 patients met the criteria for the diagnosis of C pneumoniae pneumonia (Table 1). Thirty patients, 24 of whom underwent thin-section CT, had evidence of C pneumoniae only. In the remaining 18 patients, evidence of infection from other pathogens, including Enterobacter species in one patient, Staphylococcus pneumoniae in two patients, both Staphylococcus pneumoniae and mycobacterium avium-intracellulare complex in one patient, S pneumoniae in seven patients, Klebsiella pneumoniae in two patients, Haemophilus influenzae in four patients, and M pneumoniae in one patient, was found. Patients with mixed infection were excluded from our study because we could not determine which pathogen was responsible for each CT finding. Thus, 24 thin-section CT scans in 24 patients (17 men, seven women; age range, 19–89 years; mean age, 70 years) with pure C pneumoniae infection were included in our study. For comparison, 41 patients (28 men, 13 women; age range, 19–91 years; mean age, 60 years) with S pneumoniae pneumonia and 30 patients (20 men, 10 women; age range, 16–67 years; mean age, 33 years) with M pneumoniae pneumonia who had no evidence of other causative pathogens and who underwent thin-section CT during the same period were also included in this study.
Thin-Section CT Techniques
A helical CT scanner (HiSpeed Advantage LX/i or HiSpeed Advantage SG; GE Medical Systems, Milwaukee, Wis) was used for all examinations, which were performed within 1 day before or after other clinical examinations. After helical CT scans of the entire lung were obtained by using a 7-mm section thickness, a 1-mm section thickness was used to obtain additional scans in those patients who demonstrated abnormal areas of opacity. These 1-mm sections were then reconstructed by using a high-spatial-frequency algorithm with a peak voltage of 120–140 kV and a current of 180–230 mA, and 0.8 or 1.0 second per section. Section intervals were variable, ranging from 1 mm (ie, no gap) to 1 cm. Targeted reconstruction was occasionally performed when the disease was restricted to a lung. Contrast material was typically not used except in those patients who were suspected of having a lung abscess or possible malignancy. The hard-copy images were displayed with a window level of –700 HU and a window width of 1500 HU.
Evaluation of Thin-Section CT Findings
Two chest radiologists (A.N. and A.S., with 12 and 7 years experience, respectively) who were blinded to the diagnoses retrospectively and independently assessed the presence of consolidation, ground-glass opacity (GGO), bronchovascular bundle thickening, nodules, pleural effusion, lymphadenopathy, reticular or linear opacity, airway dilatation, and pulmonary emphysema on 95 thin-section CT scans of the three kinds of pneumonia. Readers also evaluated if the pneumonia was unilateral or bilateral and identified the opacity pattern of pneumonia. In addition, readers recorded which lobe of the lung was involved.
Consolidation was defined as airspace opacification with obscuration of the underlying vasculature. Thickening of the bronchovascular bundle was defined as an apparent thickening of the bronchovascular bundle in comparison with the unaffected lung parenchyma. Nodules that were present in the bronchovascular bundle were regarded as peribronchovascular nodules. Bronchial wall thickening was defined as bronchovascular bundle thickening because this finding was indistinguishable from peribronchial interstitial thickening or alveolar infiltration. There were often equivocal thin-section CT appearances that could be described as either consolidation or bronchovascular bundle thickening. For simplification and maintenance of reproducibility for CT interpretation, we regarded all the areas of peribronchovascular opacity as bronchovascular bundle thickening as long as these areas assumed a branching shape, which was suggestive of a bronchovascular distribution of the disease. GGO was defined as mildly increased attenuation without obscuration of the underlying vasculature. We ignored minimal GGO at the periphery of consolidation and other areas of opacity because such findings can be seen in almost all cases of pneumonia. Nodules were divided into three types—that is, centrilobular nodules, which were defined as small nodules in a centrilobular location; peribronchovascular nodules, which were defined as relatively larger nodules that were associated with the bronchovascular bundles; and other nodules, which were defined as nodules that were not associated with centrilobular structures or bronchovascular bundles. These other nodules might be only a few centimeters in diameter, which is suggestive of granulomas. If more than one type of nodule was seen, the most predominant type was represented.
Mediastinal lymphadenopathy was judged to be present when the minimal diameter of a lymph node was larger than 10 mm. Hilar lymphadenopathy was judged to be present only if the maximum diameter of the ipsilateral hilum exceeded that of the contralateral hilum by 1.5-fold or more. For the assessment of pleural effusion and lymphadenopathy, we used a mediastinal window of 7 mm for helical CT scans obtained with the standard algorithm because the mediastinal windows of thin-section CT with the bone algorithm were noisy and inadequate for evaluation of mediastinal structures or pleural effusion.
Interlobular septa thickening, intralobular interstitial thickening, and areas of irregular linear opacity were all classified as reticular or linear opacity. The reticular framework in GGO that is described as crazy-paving appearance was not classified as an area of reticular or linear opacity because this framework has been shown to represent part of alveolar opacification in about two-thirds of cases (10).
Airway dilatation was considered to be present if the bronchial caliber exceeded that of the adjacent pulmonary artery.
Pulmonary emphysema was defined as scattered or diffuse areas of low attenuation without internal vascular structure in comparison with normal lung parenchyma.
The 95 cases of pneumonia were also classified according to the predominant type of opacity seen at thin-section CT—that is, consolidation predominance; bronchovascular bundle thickening predominance, which included peribronchovascular nodules and centrilobular nodules; and GGO predominance. These three categories were consistent with the classically known airspace pneumonia, bronchopneumonia, and interstitial pneumonia.
The final decision for the presence of each finding and the opacity pattern for each case was reached by the consensus of the two readers.
Statistical Analysis
A commercially available software program (SPSS, version 11.0.1; SPSS, Chicago, Ill) was used for statistical analysis. For each finding, 
values were calculated between the readers. The frequency of the findings, including the prevalence of bilateral pneumonia, was compared statistically among the three types of pneumonia by using the 2 or Fisher exact test. In addition, the frequency of bronchovascular bundle thickening and peribronchovascular or centrilobular nodules, which was an indicator of bronchopneumonia, was compared among the three types of pneumonia by using the 2 or Fisher exact test. The average age of the patients was also compared among the three groups by using a Student t test.
We compared the frequency of the three types of predominant opacity patterns and the involvement of each lobe among the three kinds of pneumonia by using the 2 or Fisher exact test. A P value of less than .05 was considered to indicate a statistically significant difference.
|
RESULTS
|
|---|
The value between readers was 0.549 for consolidation, 0.534 for GGO, 0.430 for bronchovascular bundle thickening, 0.721 for nodules, 0.832 for pleural effusion, 0.642 for lymphadenopathy, 0.511 for reticular or linear opacity, 0.587 for airway dilatation, 0.697 for pulmonary emphysema, 0.817 for bilateral lung involvement, and 0.594 for pneumonia type. These values indicated fair to good interreader agreement.
Consolidation, nodules, and bronchovascular bundle thickening were the most common findings in patients with C pneumoniae pneumonia (Table 2, Figs 1, 2). Other nodules that were suggestive of granuloma, however, were seen in four patients (17%). Areas of reticular or linear opacity were often associated with C pneumoniae pneumonia. GGO, bilateral lung involvement, and pulmonary emphysema were seen in about half of the patients with C pneumoniae pneumonia (Fig 3). The frequency of involvement in each lobe (P = .330) and the frequency of involvement in more than one lobe (P = .185) were not statistically different between the three types of pneumonia (Table 3).
View larger version (130K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 1: Transverse thin-section CT scan demonstrates C pneumoniae bronchopneumonia with airway dilatation in 66-year-old man. Centrilobular nodules (arrows) and lobular areas of consolidation with bilateral airway dilatation (arrowheads) are seen. Note the associated areas of linear opacity.
|
|
View larger version (138K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 2: Transverse thin-section CT scan demonstrates C pneumoniae pneumonia with consolidation predominance in 79-year-old man. Segmental consolidation with partially intermingled areas of GGO in right middle lobe and patchy areas of GGO in both lower lobes (*) are seen. Associated linear opacities (arrows) and scattered dotlike areas of low attenuation (arrowheads), which are indicative of centrilobular emphysema, are also noted.
|
|
View larger version (95K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 3: Transverse thin-section CT scan demonstrates C pneumoniae pneumonia with GGO predominance in 23-year-old man. Extensive areas of GGO are seen in the left lower lobe, which demonstrates thickened bronchovascular bundles (arrows) and fine reticular framework (ie, crazy-paving appearance).
|
|
Bronchovascular bundle thickening (P = .022) and airway dilatation (P = .034) were more frequently seen with C pneumoniae than with S pneumoniae (Table 4, Fig 4). Although the presence of pulmonary emphysema was higher in patients with C pneumoniae pneumonia than in those with S pneumoniae pneumonia, the difference was not statistically significant (P = .255). Bronchovascular bundle thickening or peribronchovascular or centrilobular nodules, which were considered to be a spectrum of thin-section CT findings for bronchopneumonia, were seen in 21 (88%) of 24 patients with C pneumoniae pneumonia, in 26 (63%) of 41 patients with S pneumoniae pneumonia, and in 28 (93%) of 30 patients with M pneumoniae pneumonia (Fig 5). The difference in frequency of bronchovascular bundle thickening and peribronchovascular or centrilobular nodules between C pneumoniae and S pneumoniae (P = .036) and between S pneumoniae and M pneumoniae (P = .004) was statistically significant. The prevalence of C pneumoniae pneumonia and M pneumoniae pneumonia, however, was not statistically different (P = .462).
View larger version (104K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 4: Transverse thin-section CT scan demonstrates S pneumoniae airspace pneumonia in 54-year-old man. Segmental consolidation with air bronchograms (arrows) is seen in right middle lobe. Focal area of GGO (arrowhead) is noted in right lower lobe.
|
|
View larger version (139K):
[in this window]
[in a new window]
[Download PPT slide]
|
Figure 5: Transverse thin-section CT scan demonstrates M pneumoniae bronchopneumonia in 23-year-old man. Branching centrilobular nodules (tree-in-bud appearance, arrowheads) are seen on a background of faint GGO. Bronchial wall thickening (arrow) is also noted.
|
|
By examining average age, we determined that patients with C pneumoniae pneumonia were significantly older than those with M pneumoniae pneumonia (P < .001). Reticular or linear opacity (P = .017), airway dilatation (P = .016), and pulmonary emphysema (P = .003) were more commonly seen in patients with C pneumoniae pneumonia than in those with M pneumoniae pneumonia.
The frequency of the three predominant opacity patterns was similar for each of the three types of pneumonia, with a mildly higher frequency of consolidation in patients with S pneumoniae pneumonia (Table 5). There was no statistically significant difference in the frequency of each opacity pattern between patients with C pneumoniae pneumonia and those with M pneumoniae pneumonia (P = .67–.95) or between patients with C pneumoniae pneumonia and those with S pneumoniae pneumonia (P = .10–.84) (Table 6).
|
DISCUSSION
|
|---|
Our results demonstrate that C pneumoniae pneumonia has a wide spectrum of thin-section CT findings that consist of consolidation, GGO, bronchovascular bundle thickening, and nodules in a different proportion. Of these findings, the presence of nodules was a remarkably common finding and was usually associated with other findings. If bronchovascular bundle thickening and peribronchovascular or centrilobular nodules were grouped together as indicators of bronchopneumonia, then the prevalence of bronchovascular bundle thickening and peribronchovascular or centrilobular nodules increased, even though these findings did not always predominate.
We also found that pulmonary emphysema and airway dilatation were frequently seen in patients with C pneumoniae pneumonia. It has been shown that chronic obstructive pulmonary disease is a predisposing factor for C pneumoniae pneumonia (11–13). We confirmed this fact by means of radiologic examination. It must be noted, however, that the prevalence of pulmonary emphysema detected at thin-section CT would be higher than that of clinically diagnosed pulmonary emphysema because thin-section CT can demonstrate areas of minimal low attenuation. We may also speculate that C pneumoniae, which occasionally results in recurrent pneumonia in patients with chronic infection, could be responsible for airway dilatation. The results of several reports suggest that current infection with C pneumoniae might cause wheezing or acute exacerbation of asthma (14,15). These observations indicate that C pneumoniae can act as a bronchial irritant. We think that prolonged or repeated irritation would lead to irreversible airway dilatation (ie, bronchiectasis). If this speculation is true, then airway dilatation should be more commonly seen in patients with recurrent C pneumoniae pneumonia. Unfortunately, our study did not discriminate between primary and recurrent infection because immunoglobulin M testing and paired serum sampling, which are essential for the proof of primary infection, were not routinely obtained.
McConnell et al (2) reported chest radiographic findings of C pneumoniae pneumonia by dividing patients into primary and recurrent infection groups. These researchers noted that areas of alveolar opacity were more common in patients with primary infection; those with recurrent infection had variable abnormalities, with slightly more areas of interstitial than alveolar opacity (2). No statement regarding airway dilatation is mentioned in the report. In some of our patients, however, airway dilatation at thin-section CT appeared as an area of reticulonodular opacity at chest radiography; this finding was probably consistent with the finding of interstitial opacity that is described by McConnell et al (2). We speculate that in the series by McConnell et al, researchers may have observed airway dilatation, especially in the group with recurrent infection, if thin-section CT had been performed.
Airway dilatation and bronchovascular bundle thickening were significantly more prevalent in patients with C pneumoniae pneumonia than in those with S pneumoniae pneumonia. We could not, however, establish specific findings that would enable definitive discrimination between these two types of pneumonia. Consolidation was a common finding and was often predominant in patients with S pneumoniae pneumonia. Associated centrilobular and peribronchovascular nodules, however, were not uncommon. Therefore, in patients with consolidation and bronchovascular bundle thickening, we are unable to suggest the possibility of either C pneumoniae or S pneumoniae. Nevertheless, bronchovascular bundle thickening and airway dilatation without consolidation might be more suggestive of C pneumoniae pneumonia than of S pneumoniae pneumonia. Conversely, consolidation without bronchovascular bundle thickening or nodules might be more suggestive of S pneumoniae.
In our series, the mean patient age was significantly higher in patients with C pneumoniae pneumonia than in those with M pneumoniae pneumonia. The age range for patients with pneumonia was similar to that described in previous reports; our study, however, excluded pediatric patients (2,4–7,16). Although we did not include a pediatric population, we believe that our results are applicable to adult patients. It must be noted, however, that C pneumoniae pneumonia also can occur in children, usually as a primary infection. At thin-section CT, patients with C pneumoniae pneumonia had a significantly higher frequency of ancillary findings, including reticular or linear opacity, airway dilatation, and pulmonary emphysema, than did those with M pneumoniae pneumonia. Nodules and airspace consolidation were commonly seen in both diseases without a significant difference. Although both C pneumoniae pneumonia and M pneumoniae pneumonia have been described as representatives of infectious bronchiolitis or bronchopneumonia (17), the diagnosis of C pneumoniae bronchopneumonia appeared to be relatively complicated owing to the presence of ancillary findings in comparison with M pneumoniae pneumonia. These differences may be derived from chronic or repetitious nature of C pneumoniae pneumonia, or these differences may merely represent preexistent lung disease that is not associated with C pneumoniae infection because the patients with C pneumoniae pneumonia were older. It should also be noted that airspace consolidation is common and appears as predominant areas of opacity in both C pneumoniae and M pneumoniae pneumonia. In such cases, C pneumoniae pneumonia is indistinguishable from M pneumoniae pneumonia and S pneumoniae pneumonia.
Centrilobular nodules with airway dilatation are commonly seen in cases of Mycobacterium avium–intracellulare complex infection or tuberculosis. In our study, however, C pneumoniae pneumonia involved the upper lobe in 16 patients, the middle lobe in 11 patients, and the lower lobe in 17 patients. We believe that there is no zonal predilection in cases of C pneumoniae pneumonia, as is seen in cases of M avium–intracellulare complex infection or tuberculosis (18). In addition, the presence of well-defined nodules that measure only a few centimeters in diameter and are suggestive of granuloma is relatively rare.
Our study had several limitations. First, the test for C pneumoniae infection was not 100% accurate and had a relatively low sensitivity. Therefore, it is possible that patients in our series had either S pneumoniae or M pneumoniae mixed with C pneumoniae. Conversely, C pneumoniae pneumonia in our series might have been caused by an undetected pathogen, especially a virus. Therefore, our results may have been affected to some extent by the uncertainty of the clinical diagnosis of C pneumoniae infection. Second, we did not attempt to exclude underlying diseases in each patient; such diseases could influence CT findings. It should be noted that airway dilatation might be a preexisting change rather than a resultant finding in patients with C pneumoniae pneumonia. Third, we investigated only those patients who had undergone CT, which was not performed in patients with slight clinical symptoms. This could be a considerable bias. Fourth, to reveal specific imaging features of C pneumoniae pneumonia, pneumonia that was caused by various other pathogens should have been included. Fifth, we did not separate C pneumoniae infection into primary and recurrent or chronic infections, which may have had different thin-section CT findings, as described earlier. Sixth, because we did not include pediatric patients, our data cannot be applied to the children who are subject to C pneumoniae infection.
In conclusion, C pneumoniae pneumonia demonstrated various thin-section CT findings. Airspace consolidation with or without centrilobular or peribronchovascular nodules is often seen in all three types of pneumonia and is nonspecific. However, centrilobular or peribronchovascular nodules or bronchovascular bundle thickening without consolidation and with pulmonary emphysema or airway dilatation in elderly patients may be more likely to be C pneumoniae pneumonia than S pneumoniae or M pneumoniae pneumonia.
|
ACKNOWLEDGMENTS
|
|---|
We are grateful to Rie Hashida, a radiology technologist from the Kofu Municipal Hospital, for the preparation of imaging data for this study.
|
FOOTNOTES
|
|---|
Abbreviations: GGO = ground-glass opacity
2 Current address: Faculty of Healthcare, Tokyo Healthcare University, Japan 
Authors stated no financial relationship to disclose.
Author contributions: Guarantors of integrity of entire study, T.A., K.O.; study concepts, A.N., K.O.; study design, A.N.; literature research, A.N.; clinical studies, A.N., K.O., M.A., Z.O., H.Y.; data acquisition, A.N., A.S., M.A., Z.O., H.Y.; data analysis/interpretation, A.N., A.S., Y.H.; statistical analysis, A.N., Y.H.; manuscript preparation, A.N.; manuscript definition of intellectual content, T.A.; manuscript editing, A.N., T.A.; manuscript revision/review, T.A., Y.H.; manuscript final version approval, T.A.
|
References
|
|---|
- Grayston JT, Diwan VK, Cooney M, Wang S. Community- and hospital-acquired pneumonia associated with chlamydia TWAR infection demonstrated serologically. Arch Intern Med 1989;149:169–173.[Abstract/Free Full Text]
- McConnell CT Jr, Plouffe JF, File TM, et al. Radiographic appearance of chlamydia pneumoniae (TWAR strain) respiratory infection. Radiology 1994;192:819–824.[Abstract/Free Full Text]
- Kuo CC, Jackson LA, Campbell LA, Grayston JT. Chlamydia pneumoniae (TWAR). Clin Microbiol Rev 1995;8:451–461.[Abstract]
- Lieberman D, Ben-Yaakov M, Lazarovich Z, et al. Chlamydia pneumoniae community-acquired pneumonia: a review of 62 hospitalized adult patients. Infection 1996;24:109–114.[CrossRef][Medline]
- Ali NJ, Sillis M, Andrews BE, Jenkins PF, Harrison BD. The clinical spectrum and diagnosis of mycoplasma pneumoniae infection. Q J Med 1986;58:241–251.[Medline]
- Itoh I, Ishida T, Hashimoto T, et al. Chest radiograph of atypical pneumonia: comparison among chlamydia pneumoniae pneumonia, ornithosis, and mycoplasma pneumoniae pneumonia [in Japanese]. Kansenshogaku Zasshi 2000;74:954–960.[Medline]
- Kauppinen MT, Lahde S, Syrjala H. Roentgenographic findings of pneumonia caused by chlamydia pneumoniae: a comparison with streptococcus pneumonia. Arch Intern Med 1996;156:1851–1856.[Abstract/Free Full Text]
- Kishimoto T, Kubota Y, Matsushima T, et al. Assay of specific anti-chlamydia pneumoniae antibodies by ELISA method. I. Evaluation of ELISA kit using outer membrane complex [in Japanese]. Kansenshogaku Zasshi 1996;70:821–829.
- Kishimoto T, Matsushima T, Morikawa T, Kawagoe K. Assay of specific anti-chlamydia pneumoniae antibodies by ELISA method. III. Setting of serological criteria [in Japanese]. Kansenshogaku Zasshi 1999;73:457–466.
- Johkoh T, Itoh H, Muller NL, et al. Crazy-paving appearance at thin-section CT: spectrum of disease and pathologic findings. Radiology 1999;211:155–160.[Abstract/Free Full Text]
- Miyashita N, Matsumoto A, Kubota Y, Nakajima M, Niki Y, Matsushima T. Continuous isolation and characterization of chlamydia pneumoniae from a patient with diffuse panbronchiolitis. Microbiol Immunol 1996;40:547–552.[Medline]
- Von Hertzen L, Alakarppa H, Koskinen R, et al. Chlamydia pneumoniae infection in patients with chronic obstructive pulmonary disease. Epidemiol Infect 1997;118:155–164.[CrossRef][Medline]
- Miyashita N, Niki Y, Nakajima M, Kawane H, Matsushima T. Chlamydia pneumoniae infection in patients with diffuse panbronchiolitis and COPD. Chest 1998;114:969–971.[Abstract/Free Full Text]
- Hahn DL, Dodge RW, Golubjatnikov R. Association of Chlamydia pneumoniae (strain TWAR) infection with wheezing, asthmatic bronchitis, and adult-onset asthma. JAMA 1991;266:225–230.[Abstract/Free Full Text]
- Miyashita N, Kubota Y, Nakajima M, Kiki Y, Kawane H, Matsushima T. Chlamydia pneumoniae and exacerbations of asthma in adults. Ann Allergy Asthma Immunol 1998;80:405–409.[Medline]
- Reittner P, Muller NL, Heyneman L, et al. Mycoplasma pneumoniae pneumonia: radiographic and high-resolution CT features in 28 patients. AJR Am J Roentgenol 2000;174:37–41.[Abstract/Free Full Text]
- Webb WR, Muller NL, Naidich DP. Airway diseases. In: Langlois A, et al, eds. High-resolution CT of the lung. 3rd ed. Philadelphia, Pa: Lippincott Williams & Wilkins, 2001; 467–546.
- Webb WR, Muller NL, Naidich DP. Diseases characterized primarily by linear and reticular opacities. In: Langlois A, et al, eds. High-resolution CT of the lung. 3rd ed. Philadelphia, Pa: Lippincott Williams & Wilkins, 2001; 193–353.
TOP
ABSTRACT
INTRODUCTION
