Sonographic Diagnosis and Follow-Up of Pneumonia: A Prospective StudyReissig A. · Kroegel C.
Pneumology and Allergology/Immunology, Department I, Medical University Clinics, Friedrich Schiller University, Jena, Germany
Background: Although transthoracic ultrasound offers several important advantages as diagnostic imaging technique in pleural and pulmonary conditions, its significance for the diagnosis and monitoring of pneumonia has yet to be established. Objectives: To identify sonographic features associated with pneumonia at admission and during the course of the disease under treatment. Methods: Thirty patients (12 females, 18 males; median age 65.5 years) with X-ray-proven pneumonia underwent transthoracic sonography (TS) on day 0, between days 1 and 3, 4 and 7, 8 and 14, 15 and 21, and after day 21. TS was assessed according to: number, location, shape, echogenicity, echotexture, echostructure, breath-depending movement, size of pneumonic lesions, bronchoaerogram, fluid bronchogram, superficial fluid alveologram, necrotic areas, vascularity and incidence of local and/or basal pleural effusion. Results:Thirty-three pneumonic infiltrates were eligible for analysis in 30 patients. In 57% (17/30), the likely pathogenic microorganism was identified. Pneumonia was recognized as a hypoechoic area of varying size (mean size between 33.7 × 9.38 and 91.2 × 45.3 mm) and shape, with irregular and blurred margins along with a nonhomogeneous echotexture. The most characteristic feature was a positive bronchoaerogram (32/33). Sixty-one percent (20/33) revealed basal and 9% (3/33) local effusion. During follow-up, lesions decreased in size or disappeared (30/33) or decreased in number (4/9). The bronchoaerogram became less pronounced (13/32), basal pleural effusion either diminished (7/20) or dissipated (7/20), as did localized effusion (3/3). In 30 cases, the course of pneumonia was comparable using X-ray and TS. Conclusions: TS is a noninvasive technique for the diagnosis and follow-up of patients with pneumonia.
Copyright © 2007 S. Karger AG, Basel
The diagnosis of pneumonia is based mainly on the patient’s medical history and physical examination. However, in severe cases further evaluation should include posteroanterior and lateral chest radiographs, oximetry, a complete blood count, serum electrolytes, renal function tests, and, if available, expectorated sputum for Gram stain and culture. In most cases, the diagnosis of pneumonia is confirmed with a chest X-ray. The use of frontal and lateral views is standard practice in chest radiography.
More recently, sonography of the lung has been used in the diagnosis of pulmonary embolism and pneumothorax [1,2,3,4]. However, little is known  about whether it can also be used in the diagnosis and follow-up of pneumonia.
The rationale underlying the present study was, therefore, to determine whether pneumonic lesions are detectable using transthoracic sonography (TS) of the chest. In addition, the study aimed at characterizing the sonomorphology of pneumonia in adults at the time of first examination and during the course of disease. Moreover, the study assessed whether specific morphometric features corresponded to the type of infectious agent causing pneumonia.
Material and Methods
A total of 82 patients with suspected pneumonia were enrolled in this prospective study. Of these, 52 patients were excluded due to the unavailability of X-ray or ultrasound control of the chest or because no pneumonia had been established by X-ray. Six of the 82 patients (7.3%) did not reveal subpleural sonographic alterations and had thus, to be excluded, too.
Thirty patients with a total of 33 pneumonic infiltrates (12 females, 18 males; median age 65.5 years, range 31–81 years) who fulfilled the inclusion criteria of clinically, X-ray- and ultrasound-established pneumonia, underwent X-ray and ultrasound control. TS was performed on day 0 (n = 33), between days 1 and 3 (n = 12), 4 and 7 (n = 23), 8 and 14 (n = 24), 15 and 21 (n = 8), and, if possible, after day 21 (n = 17). Results from TS were compared to chest X-ray data. Day 0 was defined as the day of the first ultrasound examination.
Of the patients, 66% had fever (20/30), 87% dyspnea (26/30), 90% cough (27/30) and 90% expectoration of purulent sputum (27/30). In 28 of 30 patients (93%), rales could be detected on auscultation. Erythrocyte sedimentation rate and C-reactive protein were elevated above normal values in all patients, whereas total white blood cell count was normal in 20 patients (66%), increased (above 11.3 × 109/l) in 8 (27%) and decreased below normal values (4.4 ×109/l) in 2 patients (7%).
Among the 30 patients, community-acquired pneumonia was diagnosed in 29 patients, 17 of whom were treated with antibiotics prior to admission. Due to treatment failure (no improvement in clinical and radiological signs), the patients were admitted to hospital. Here, the original treatment was discontinued and diagnostic procedures (including TS) were performed before an alternative antibiotic regimen was initiated. The mean antibiotic treatment period was 14 days (range 9–65 days). The remaining 12 patients, admitted without any prehospital therapy, suffered from severe symptoms and comorbidity, respectively.
The study protocol followed the guidelines of the local ethics committee.
Sonographic investigation was conducted using a 5-MHz convex scanner, occasionally supplemented by examination with a 7.5-MHz linear scanner or color flow Doppler mode (AU5 Harmonic; Esaote Biomedica, Munich, Germany). The patients were examined posteriorly in a seated and a lying position from the anterior aspect of the chest. Sonography entailed a systematic examination of all intercostal spaces of both sides of the lung. Selected sonographic pictures of pneumonic lesions were documented on photographic film.
The following criteria were assessed employing TS: (1) number, (2) location, (3) shape, (4) echogenicity, echotexture and echostructure, (5) breath-depending movement and (6) size (including length, height, circumference and surface) of pneumonia. Furthermore, incidences of (7) bronchoaerogram, (8) fluid bronchogram, (9) superficial fluid alveologram, (10) necrotic areas, (11) vascularity and (12) local and/or basal pleural effusion were reviewed. For a definition of the main diagnostic sonomorphologic criteria of TS, see our previous studies [6,7,8].
The main diagnostic sonomorphologic criteria of TS are defined as follows . The bronchoaerogram impresses as multiple small air inlets within a consolidation measuring a few millimeters in diameter or as a tree-shaped echogenic structure.
A fluid bronchogram represents exudate-packed conducting airways. It occurs less frequently than the bronchoaerogram, is characterized by echo-free tubular structures along the airways and can be differentiated from pulmonary vessels using color Doppler imaging. The fluid bronchogram reflects exudate-packed conducting airways and may indicate a poststenotic pneumonia.
A superficial fluid alveologram is defined as a small and homogeneous subpleural section without air or fluid bronchograms.
Local pleural effusion is characterized by a very small echo-poor/echo-free space between the pleura visceralis and parietalis. In this study, local effusion reflects pleural effusion adjacent to the pneumonic infiltration.
Basal pleural effusion is demonstrated as an echo-poor/echo-free space between the pleura visceralis and parietalis, best visible above the diaphragm.
All patients underwent standard posteroanterior chest radiography (Vertix Polydros SX 80, Siemens, Erlangen, Germany, or Digitaldiagnost, Philips, Hamburg, Germany). In 25 patients, an additional chest X-ray in a side plane was performed. X-rays were analyzed by experts in chest radiology.
Thirty-three pneumonic lesions in 30 patients were evaluated by TS. Pneumonia-associated alterations of the lung parenchyma are observed as an echo-poor, essentially nonhomogeneous area with irregular and blurred margins. In addition, the lesions varied in shape and size with positive bronchoaerogram and a superficial fluid alveologram. The dimension of the lesions ranged between 33.7 × 9.38 and 91.2 × 45.3 mm. Multiple hyperechoic inclusions indicating air inlets within the pneumonic lesions (positive bronchoaerogram) were shown in 32 of 33 pneumonias. Intense bronchoaerograms were visible in 3 of the 32 cases, whereas the bronchoaerogram in 6 patients was weak (i.e. only few hyperechoic inclusions were found). No bronchoaerogram could be detected in only 1 patient (table 1). None of the patients revealed necrotic areas within the pneumonic lesion.
|Table 1. Sonomorphology of patients with X-ray-proven pneumonia on day 0 (33 pneumonias in 30 patients)|
Assessment of vascularity of the pneumonic lesions was performed in 11 patients. In 7 cases, no vessels could be detected, whereas in 4 cases vessels could be identified. Three vessels were located in the periphery. One of them was identified as a bronchial artery by color Doppler sonography. In the fourth patient, vessels were detected both at the periphery and in the centre of the lesion, most likely reflecting bronchial and pulmonary arteries.
Twenty of the 33 pneumonias (61%) were associated with an uncomplicated basal effusion. Three of the 33 pneumonic lesions (9%) revealed a local accumulation of pleural fluid corresponding to the pneumonic infiltrate.
Lesions were distributed equally, although slightly more pneumonic infiltrates were detected on the right side (15 versus 12). In 3 cases, pneumonic lesions were observed on both sides of the lung (table 1). The sonomorphologic features of pneumonia on day 0 are shown in table 1.
Patients were followed-up for more than 21 days in order to obtain more information on the sonomorphology of parenchymal lesions during the course of pneumonia.
In 91% (30/33) of the pneumonic lesions, a decrease or disappearance of the hypoechoic, irregularly shaped parenchymal lesion, as measured by length and height as well as circumference and area of infiltration, could be observed (fig. 1). In 2 cases (6%), however, pneumonic infiltration initially increased in size at the time of the first follow-up examination, before a reduction in size was observed thereafter. In 1 patient who developed chronic pneumonia, no significant change in size could be observed over the study period of 21 days. Among the 9 patients with multiple pneumonic lesions, a reduction in the number of lesions could be detected in 4 cases, whereas in the remaining patients the number of pneumonic infiltrates remained constant. In addition, the echo representing air inclusions became less intense in 13 of 32 cases (41%). In the remaining 19 patients, the intensity remained unchanged during follow-up. Local pleural fluid corresponding to the pneumonic infiltrate was observed in 3 cases, but disappeared in all cases during follow-up. Among the 20 patients with basal pleural effusions, the extent of the pleural fluid accumulation remained unchanged in 6 cases, decreased in 7 (in 2 cases basal effusion changed into localized effusion) and disappeared in the remaining 7 patients (fig. 2).
|Fig. 1. Sonographically measured area of 33 pneumonic infiltrates in 30 patients during the course of disease. The bold line indicates the median size of the pneumonic infiltration. *Median period of antibiotic treatment.|
|Fig. 2. Sonomorphologic characterization of pneumonic lesions during the course of the disease under antibiotic treatment.|
In 30 cases, the course of pneumonia was identical using both X-ray and TS. In the majority (27 cases), a continuous regression of the pneumonic infiltrate was observed by TS as well as by X-ray (fig. 3). In 2 cases, an initial increase in pneumonic lesion size was observed which decreased thereafter, as shown by TS and chest radiography. In 1 of the patients, a new pleural effusion appeared along with an increasing pneumonic lesion. Discrepancies between TS and X-ray were detected in 3 cases (9%), where, although sonographic features had completely resolved, X-ray still showed small residua (fig. 4).
|Fig. 3. Sonograms and corresponding X-rays during follow-up of a 73-year-old female patient suffering from pneumonia in the left lower lobe caused by Coxiella burnetii. Sonogram A shows a bowl-shaped hypoechoic area with blurred margins and positive bronchoaerogram as well as a positive superficial fluid alveologram. The lesion measured 66.3 × 34.4 mm (circumference 16.6 cm, infiltrated area 16.5 cm2) on day 0. A small basal pleural effusion could be shown on day 0. Corresponding X-ray verifies a large pneumonic infiltrate in the left lower lobe (B). Sonogram on day 7 (C) demonstrates a pneumonic lesion with a less pronounced bronchoaerogram measuring 16.1 × 13.7 mm (circumference 0.13 cm, infiltrated area 1.7 cm2) without any basal pleural effusion. Chest radiography (D) on day 10 reveals regression of pneumonic lesions, as does the sonogram on day 11 (E). The pneumonic lesion is now wedge shaped and no bronchoaerogram is visible. Pneumonia is further reduced in size, measuring 11.7 × 16.3 mm (circumference 0.47 cm, infiltrated area 1.4 cm2). Sonogram (F) and X-ray (G) taken on day 28 detect no pathological finding.|
|Fig. 4. Sonograms and corresponding X-rays during follow-up of a 62-year-old female patient with pneumonia in the left lower lobe caused by Legionella. Sonogram A shows a hypoechoic area with blurred margins and positive bronchoaerogram and a positive superficial fluid alveologram. The lesion measured 67.4 × 25.3 mm (circumference 16.2 cm, infiltrated area 11.9 cm2) on day 0. A basal pleural effusion was also seen on day 0. Corresponding X-ray (B) reveals pneumonia in the left lower lobe. Sonogram C demonstrates a pneumonic lesion on day 7 with a less pronounced bronchoaerogram, measuring 59.1 mm × 19.4 mm (circumference 13.0 cm, infiltrated area 8.0 cm2). The basal pleural effusion was also reduced. On day 14, sonogram D shows pneumonia further reduced in size, measuring 42.3 × 8.33 mm (circumference 0.95 cm, infiltrated area 3.5 cm2). The basal effusion had completely resolved on day 14. Corresponding X-ray (E) on day 14 demonstrates regression of pneumonic infiltrates. On day 27, sonographic features (F) had completely resolved, while small residua could still be detected on chest X-ray (G).|
Pathogens causing pneumonia were identified by microbiologic evaluation of either the expectorated sputum or the bronchial secretion, using blood culture as well as serology. In 17 of 30 patients (57%), the likely pathogen responsible for pneumonia could be cultured, while in the remaining patients no microorganism was identified (table 2). Most of the patients (29 of 30) had community-acquired pneumonia and only 1 could be classified as hospital-acquired pneumonia (table 2). In 21 (70%) of the patients, pneumonia was associated with comorbidity, mostly with chronic obstructive pulmonary disease (n = 16). Among the patients with identified pathogen, 3 revealed weak bronchoaerograms and 1 showed no positive bronchoaerogram. However, no other significant sonomorphologic feature indicative of a specific pathogen was identified.
|Table 2. Clinical characteristics of the 30 adult patients with pneumonia studied|
In the present study, novel evidence demonstrating that pneumonic infiltration suspected by clinical presentation and confirmed by chest X-ray is associated with lung lesions assessable by TS imaging is presented. Typical sonomorphologic features include a hypoechoic area with irregular and serrated margins that exhibits heterogeneous echotexture caused by focal air inclusions, a small homogeneous subpleural section without air inlets, a widening of the pleural space adjacent to the pneumonic lung tissue and a gross basal pleural effusion accumulating in the costophrenic angle. On follow-up, parenchymal pneumonic lesions decreased both in size and number, lost their echotexture intensity or even disappeared over a period of more than 21 days. These data demonstrate that, in addition to chest radiography and computed tomography (CT) of the chest, TS represents a further imaging technique for identifying pneumonic lesions, and thus provides an additional tool for diagnosing pneumonia.
Positive bronchoaerogram is a common radiographic finding in conventional radiograms . The development of this phenomenon essentially requires the replacement of the alveolar lung tissue around the bronchi with fluid or soft tissue density and most often indicates alveolar disease. Therefore, the multiple air inlets within the hypoechogenic area suggest that the lesion detected by TS indeed represents the alveolar lung infiltration typically observed in pneumonia.
Although most patients fail to show parapneumonic pleural effusions, up to 20–60% of the patients hospitalized for pneumonia have radiographic evidence of pleural fluid [10,11,12]. In agreement with these data, basal pleural effusion was observed in 61% (20/33) of the patients in this study. In addition, localized pleural effusion was seen in 9% (3/33) of the infiltrates identified. Most of the parapneumonic effusions are uncomplicated and will resolve with antibiotic therapy alone. However, 5–10% of the hospitalized patients with pneumonia follow a complicated course and require a pleural drainage for full recovery [13,14,15,16]. In the present study, typical complications associated with pneumonia were not discerned, probably due to the limited number of patients.
As pointed out above, sonographic imaging is able to show a variety of features typical of bacterial pneumonia that are already well known from chest X-ray including parenchymal infiltration, air inclusions and pleural involvement. In contrast to these similarities between conventional radiography and sonography of the chest, there is also a number of differences between the 2 imaging techniques. Firstly, while on chest X-ray a bronchoaerogram features negative shadows against the more opaque pulmonary parenchyma, in sonography air inclusions are characterized by hyperechogenic areas. Secondly, the extension of the infiltrate can be estimated on conventional chest radiography, providing an additional parameter for monitoring treatment response. Although the extension of the infiltrated area was measured in the present study, the true size of the pneumonic infiltration could not be accurately assessed using sonography due to the following reason. Distal of the solid lesions, an area of increased echogenicity with multiple amplification artifacts is usually visible. Although this represents the distance which can be penetrated by ultrasound waves, it does not equal the distal border of the infiltrate, since propagation of the ultrasound waves is abruptly stopped by air inclusions. Thus, ultrasound appearance of pneumonic lesions most likely underestimates the true extension of the pneumonic infiltrate. While this fact precludes an assessment of accurate size, the data presented herein demonstrate that it does not diminish the value of TS in diagnosing and monitoring the course of pneumonia. The third difference between X-ray and sonography relates to motion artifacts. In conventional imaging techniques, such as X-ray or CT of the chest, motion artifacts pose a significant problem in obtaining an optimal image. For both, conventional radiography as well as CT, motion artifacts usually deteriorate image quality. In contrast, real-time assessment in sonography offers additional information including the breath-depending motion of the lesions during real-time investigation which may help identify the location of the parenchymal alteration and determine the extension of the process and its relation to the pleura.
Pneumonic infiltrates were found equally distributed between the two sides of the lung (12 left versus 15 right). In the majority of cases (27/30, 90%), only one side was affected, whereas in 10% (3/30) infiltrations were observed on both sides. Also, a single pneumonic lesion was observed in 24/33 (73%) cases, but interestingly, more than 1 parenchymal infiltration was found in 9/33 (27%) patients. This is surprising since multiple pneumonic infiltrates are only rarely detected on conventional radiography in otherwise healthy subjects. However, 2 potential reasons may explain this finding. First, because 70% of our patients had an underlying pulmonary disease (table 2), comorbidity may have contributed to an increased frequency of multiple infiltrated areas. Second, in chest X-ray standard projection gives a summation image resulting from superimposed normal and abnormal or partially affected lobules. In contrast, sonography allows examination along the circumference of the lung which may differentiate between single affected parenchymal sections. However, sonography cannot identify whether infiltrates converge in more distal parenchymal areas.
A number of limitations in our analysis should be noted. The major study aim was to evaluate the potential of TS imaging in depicting acute pneumonic lesions of patients with established pneumonia. The major inclusion criteria used in this study was the demonstration of pneumonic lesions in a conventional plain X-ray of the chest precluding comparison of TS with the established radiographic technique. Thus, the sensitivity and specificity of TS imaging compared with chest X-ray in detecting early pneumonia could not be established. Therefore, the results of our study need to be confirmed by prospective studies with a larger number of patients. Finally, as some of the pneumonic lesions may not extend to the pleural surface of the lung, centrally located pneumonic infiltrates are not assessable by sonography. Yet, the results presented indicate that pneumonia can be recognized in about 90% of the patients, which is in good agreement with a previously published report . Therefore, inconspicuous sonographic images in patients with suspected pneumonia cannot fully exclude the condition.
The results presented above demonstrate that TS is an additional imaging technique applicable for the diagnostic confirmation in patients with suspected pneumonia. As sonography is noninvasive, the method may be especially useful in the diagnosis of outpatients, children and pregnant women in whom radiation is either not available or should be avoided. In addition, the results also show that sonography can be used to monitor patients with pneumonia as a basis for further diagnostic and treatment-related decisions. This information can help differentiate pneumonia from other pulmonary diseases and guide subsequent treatment rationales and is useful for optimizing the most appropriate treatment so as to minimize radiation exposure, morbidity and period of convalescence. The simple use and wide availability of sonography will likely encourage more frequent patient reevaluations, new approaches to the recognition and management of acute exacerbations of pneumonia, and enhance the understanding of the natural history of pneumonia. Thus, X-ray control may only be indicated if sonography shows complete recovery or if any complications are detectable.
The authors are indebted to N. Kroegel, B.Sc., for reviewing the manuscript.
We are grateful to the Institute of Diagnostic and Interventional Radiology, Friedrich Schiller University (director: Prof. Dr. med. Dipl. chem. W.A. Kaiser) for evaluating chest radiographs and CT scans.
Angelika Reissig, MD
Allergology/Immunology, Medical Clinic I
Friedrich Schiller University, Erlanger Allee 101, DE–07740 Jena (Germany)
Tel. +49 3641 932 4131, Fax +49 3641 932 4132
Received: July 3, 2006
Accepted after revision: December 13, 2006
Published online: February 27, 2007
Number of Print Pages : 11
Number of Figures : 4, Number of Tables : 2, Number of References : 16
Respiration (International Journal of Thoracic Medicine)
Vol. 74, No. 5, Year 2007 (Cover Date: August 2007)
Journal Editor: Bolliger, C.T. (Cape Town)
ISSN: 0025–7931 (print), 1423–0356 (Online)
For additional information: http://www.karger.com/RES