m JOtlll~

oF

UL'U'~LtI~UND
ELSEVIER European Journal of Ultrasound 4 (1996) 169-176

Clinical paper
Sonographic diagnosis of pneumonia and bronchopneumonia

A. Benci a'*, M. Caremani a, D. Menchetti a, A.L. Magnolfi b

aDepartment of Infectious Diseases and Ultrasonography Unit, Ospedale Santa Maria Sopra i Ponti, Via Fonte Veneziana 1,
52100 Arezzo, Italy
bDepartment of Radiology, Ospedale Santa Maria Sopra i Ponti, Via Fonte Veneziana I, 52100 Arezzo Italy

Received 6 March 1996; revised 25 September 1996; accepted 26 September 1996

Abstract

Objectives: The usefulness of sonography in the diagnosis of pneumonia and bronchopneumonia is considered.
Ultrasound is compared with conventional radiology. Methods: Eighty patients with respiratory failure (40 of whom
were HIV positive) were randomly divided into two groups of 40. The first group was X-rayed and subsequently
subjected to an ultrasound examination of the chest. The second group was examined, first by ultrasound and
subsequently with conventional radiology. Results: The study shows that the infective lung diseases that cause alveolar
consolidation present similar, characteristic ultrasound patterns: large hypoechoic lesions or small roundish subpleural
hypoechoic lesions with fine echoes inside and occasionally with asonic canalicular formations and hypoechoic linear
structures with comet tails ('liver like' images). The diagnostic sensitivity of the ultrasound examination is comparable
to that of conventional radiology (100% vs. 90%). On the other hand, ultrasound does not detect alterations of
interstitial pneumonia, whereas conventional radiology is diagnostically effective. High resolution computed tomogra-
phy and bronchoscopy must be employed in doubtful cases. Conclusion: Ultrasound is shown to be reliable, safe and
easy to carry out, especially in immunocompromised bedridden patients, and it can play a significant role in the
diagnosis of pneumonia and bronchopneumonia. Copyright © 1996 Elsevier Science Ireland Ltd.

Keywords: Ultrasound; Pneumonia; Bronchopneumonia

Abbreviations: US, ultrasound; CR, conventional radiology; 1. Introduction

HRCT, high resolution computed tomography; PCP, pneumo-
cistis carinii pneumonia; AP, alveolar pneumonia; BP, Bron- P n e u m o n i a and b r o n c h o p n e u m o n i a continue to
chopneumonia; IP, Interstitial pneumonia; PE, pleural
effusion; NOR, normal; BAL, broncho aspiration liquid; TBB, contribute significantly to hospitalization in in-
transbronchial biopsy. dustrialized countries, a l t h o u g h the mortality rate
* Corresponding author. TeL/fax: + 39 575 305254. has declined sharply with the introduction o f an-

0929-8266/96/$15.00 Copyright © 1996 Elsevier Science Ireland Ltd. All rights reserved
PII S0929-8266(96)0019 5-4
170 A. Benci et al. / European Journal o[ Ultrasound 4 (1996) 169-176

tibiotic treatment in clinical practice. Incidence plicating diseases that might arise. This technique,
varies from 5 to 10 cases per 1000 of the popula- while showing high diagnostic sensitivity, may fail
tion per year for those between the ages of 5 and to reveal small pleural effusions and occasionally
60 (Foy et al., 1979; Jokinen et al., 1993), but does not distinguish between interstitial and alve-
there are many conditions which lead to predispo- olar forms, and between alveolar forms and other
sition, such as the ageing of the population, the pulmonary diseases. Finally, radiographic exami-
increase in immunocompromised patients, alco- nation represents a risk in certain physiological
holism, tumors, smoking, and the onset of antibi- conditions, such as infancy and pregnancy, and it
otic-resistent germs. is difficult to apply to certain bedridden patients.
Aetiologic diagnosis is possible, however, in less Chest US has been used very little in the diag-
than 50% of cases, as the micro-organism can be nosis of infectious lung diseases. Its employment
isolated from the pneumonial lesion with certainty is limited to the diagnosis and control of pleural
only through bronchial lavage, transbronchial effusions, to the study of mediastinal
biopsy and lung biopsy (Mandell, 1995). lymphadenopathies, peripheral pulmonary lesions
Pneumonias tend to conform to one of three (cysts, abscesses and tumors), and lesions of the
pathologic and ronetgenographic patterns: alve- pleura and of the chest wall, as well as to guiding
olar pneumonia (AP), bronchopneumonia (BP) the execution of bioptic and evacuative punctures
and interstitial pneumonia (IP). In AP the organ-
(Kroschel et al., 1991; Mikloweit et al., 1991;
ism causes an inflammatory exudate that spreads
Targhetta et al., 1991; Wu et al., 1991; Hsu et al.,
from one alveolus to the next, while the bronchi,
1992; Yang et al., 1992; Mathis and Difschmid,
relatively uninvolved, remain patent. The X-ray
1993).
shows a non-segmental consolidation with air
These limitations are due mainly to the air in
bronchograms. BP consists of inflammation in the
the lung which creates an impermeable barrier to
conducting airways, especially terminal and res-
US. As a result it is only possible to explore liquid
piratory bronchioles and the surrounding alveoli,
or surface parenchymatous structures in direct
sometimes casusing atelectasis, whereas airbron-
chograms are absent. BP tends to maintain a contact with the chest wall or the diaphragm and
distribution corresponding to the involved pul- without the overlapping of normally aired pul-
monary segment. Some agents cause an interstitial monary parenchyma.
pneumonia, where inflammation is predominantly Recently, studies have been carried out to eval-
in the interalveolar septa. uate the usefulness of US in the diagnosis of
Conventional radiology (CR) of the chest is infective pulmonary lesions, showing characteris-
preferred for the evaluation of patients with sus- tic US images (Targhetta et al., 1991; Mathis et
pected pneumonia and generally makes it possible al., 1992; Gehemacher et al., 1995).
to differentiate between alveolar and interstitial The purpose of our study was to determine
forms. For the latter it is sometimes necessary to whether the US examination of the chest could
apply other techniques, such as pulmonary detect the presence of an infective pneumonopa-
scintigraphy with gallium citrate 67 (Wassie et al., thy leading to alveolar consolidation, to identify
1994; Miller et al., 1995) or high resolution com- the US pattern, and to compare the diagnostic
puted tomography (HRCT) (Primack et al., sensitivity of US with that of CR.
1994), especially with AIDS patients, as in some
cases CR is unable to reveal the presence of an
interstitial pneumonia correlated with Pneumono- 2. Materials and methods
cyctis carinii (PCP) infection (Meduri and Stein,
1992). In order to evaluate the accuracy of US in the
Chest X-ray is also useful for monitoring the diagnosis of pneumonia and bronchopneumonia
course of the illness, as it can evaluate the effec- 80 patients with fever and respiratory signs, were
tiveness of the treatment, and for diagnosing com- selected, over the period September 1992 to Sep-
 A. Benci et al. / European Journal of Ultrasound 4 (1996) 169-176 171

tember 1994, from those hospitalized in the De- reflection of air (Fig. 1). This pattern can be
partment of Infectious Diseases of the Hospital of traced to small bronchial pulmonary concentra-
Arezzo. The patients were divided into two tions. The second type shows roughly triangular
groups of 40 each; in the first group (16 females hypoechoic formations from 4 to 8 cm in size with
and 24 males, average age 41 years) the patients fine echoes inside and occasionally with asonic
were subjected to chest X-rays in anteroposterior canalicular formations and hyperechoic linear
and laterolateral projections by the hospital's Ra- structures with comet tails (Fig. 2). This qiver-
diology Service. They subsequently underwent US like' US image is indicative of more extended
examinations of the chest by physicians with con- lesions (lobar and sublobar pneumonitis) which
siderable experience in US techniques. The pa- may be accompanied by canalicula structures
tients of the second group (14 females and 26 which take the form of vases and small bronchi.
males, average age 36), on the other hand, were The vessels can be differentiated from a fluid
first given chest US examinations and then chest bronchogram because of their pulsative nature,
X-rays; 22 patients of the first group and 18 of the therefore conventional ultrasound Doppler or
second were HIV positive. color Doppler can be useful.
Chest US was carried out with an Ansaldo On the basis of these findings, the US and CR
AU-560 apparatus, with a convex probe of 3.5 examination results in the two groups (Table 1)
MHz. The lungs were examined with medio-lat- show that in the first group (CR > US) 11 pa-
eral anterior and posterior intercostal imaging. tients were CR-normal and 24 US-normal. Of the
The patients were also given hemoculture, sputum 24 US-normal, 11 also appeared normal to CR,
culture and serological exams. while the other 13 belonged to the group of 15
HIV positive patients with radiological traces of cases diagnosed and confirmed as CR IP.
interstitial pneumonia were subjected to bron- The US showed a greater number of instances
choscopy with broncho lavage and to trans- of AP (16 vs. 14). This is because the US, in
bronchial biopsy (TBB). In doubtful cases the addition to confirming the 14 AP shown by the
diagnosis was completed with HRCT. CR, also pointed out two alveolar lesions in two
of the 15 cases considered interstitial after radio-
graphic examination.
3. Results In the second group (US > CR) the US initially
appeared normal in 19 patients, 10 of whom were
The radiographic diagnoses were divided sche- subsequently proven by CR to be affected by IP
matically into three types: normal (NOR), alve- and nine to be normal after a radiographic con-
olar pneumonia (AP) and interstitial pneumonia trol.
(IP). Pleural effusion (PE) was also taken into The US, however, revealed two alveolar lesions
account. more than the CR (21 vs. 19) in two patients that
The US diagnoses were subdivided into two proved NOR after the initial radiographic exami-
types: absence of alterations (NOR) and presence nation. In these two cases there were small hypoe-
of parenchymatous-like hypoechoic lesions indica- choic lesions (1.4-2 cm) located in the left basal
tive of alveolar pneumonia (AP). Anechoic forma- paracardiac region.
tions in the pleural region, indicative of pleural As regards the PE, always concomitant with
effusions (PE), were also considered. alveolar pneumonias, the US showed higher diag-
Pulmonary lesions appeared with two different nostic accuracy in the two groups (5 vs. 3),
US patterns. The first showed small subpleural though not at a significant level, given the small
hypoechoic lesions (from 1.4 to 4 cm in diameter), sample.
roundish or triangular in shape with fine echoes The four cases of diagnostic error in CR com-
inside and delimited by a hyperechoic halo which pared with US (two in the first group and two in
fades into the hyperechogenousness of the sur- the second) were reexamined by another radiolo-
rounding pulmonary parenchyma, due to total gist, who confirmed the initial diagnosis.
172 A. Benci et al. / European Journal oJ' Ultrasound 4 (1996) 169-176

Fig. 1. Ultrasonographic aspects of bronchopneumonia. (Left) A triangular hypoechoic lesion, (Right) a small, roundish hypoechoic
lesion sourronding by strong hyperechoic rim.

All the patients with a radiographic diagnosis HRCT revealed an obstructive pneumonia. In two
of IP (15 in the first group and 10 in the second) cases which with CR appeared to be resolved US
and the two patients of the second group who revealed small subpleural hypoechogenes, which
appeared CR-normal, but not US normal, were cleared up completely in the following days.
subjected to HRCT and subsequently to bron-
choscopy with BAL and TBB in an effort to
discover opportunistic pathogenes (Table 2). 4. Discussion
In 23 patients (13 in the first group and 13 in
the second) HRCT confirmed the presence of IP. Bronchopneumonias, segmental pneumonia
More importantly, it confirmed the diagnosis of and, in particular, lobar pneumonias, are charac-
AP suggested by US both in the two patients of terized by the production of a fibrino-purulent
the first group interpreted as IP by the CR and in exudate, wich occupies the small bronchial tubes
the two patients of the second group who ap- and the bronchioli as well the corresponding alve-
peared normal after radiography. oli (recall that most alveoli are distributed periph-
Finally, all patients affected by pneumonia and erally). This leads to the hypothesis that all
alveolar bronchial pneumonia were monitored pulmonary consolidations, concerning a certain
during antibiotic treatment, both with CR and number of alveoli, are detectable on the surface of
with US. The US monitor made it possible to the lung by means of chest US. Probably all
follow the progressive disappearance of the le- pneumonias, included the central ones or the
sions which gradually decreased from parenchima small ones, could be identified with US examina-
to pleural edge (Fig. 3). In the first group only tion, as the hyperechoic interphase is replaced
one case of pneumonia with homogeneous hypoe- peripheraly by a solid pattern, due to fibrino-pu-
choic pattern had no US improvement and a rulent material, occuping the subpleural alveoli.
 A. Benci et al. / European Journal of Ultrasound 4 (1996) 169-176 173

Fig. 2. A liver-like pulmonary lesion with echofree and hyperechoic structures (comet tails) due to Streptococcus pneumoniae and a
left-sided mirror artefact (arrows).

A recent s t u d y o f U S i m a g i n g o f p n e u m o n i a , (100%) (Table 3) in the d i a g n o s i s o f p n e u m o n i a

s h o w e d only 88.8% o f U S a b n o r m a l i t y in the t o t a l a n d bronchopneumonia. A c t u a l l y , C R also shows
o f p a t i e n t s with e s t a b l i s h e d r a d i o l o g i c a l p n e u m o - a high d i a g n o s t i c sensitivity (90%) b u t less t h a n
nia ( G e h e m a c h e r et al., 1995). O u r investigation,
a l t h o u g h it was c o n d u c t e d o n a limited n u m b e r o f
Table 2
patients, d e m o n s t r a t e s t h a t U S is quite sensitive Aetiological diagnosis in patients subjected to HRCT (high
resolution computed tomography) and bronchoscopy with
Table 1 BAL (bronchoscopic aspiration liquid) and TBB (trans
US and radiological findings in the two groups of patients bronchial biopsy)

GROUP Lesion CR US Number of Aetiology Diagnosis

patients
I NOR 11 24
IP 15 / 20 Pneumocystis carinii IP
AP 14 16 2 Pneumocystis carinii + IP
PE 2 3 Cytomegalovirus
II NOR 11 19 1 Cytomegalovirus IP
IP 10 / 1 Enterobacter AP
AP 19 21 1 Acinetobacter AP
PE l 2 2 Streptococcus pneumoniae AP
27 Total
NOR, normal; IP, interstitial pneumonia; AP, alveolar pneu-
monia; PE, pleural effusion. IP, interstitial polmonitis; AP, alveolar polmonitis.
174 A. Benci et al. / European Journal o f Ultrasound 4 (1996) 169-176

Fig. 3. Progressive reduction of the pulmonary lesion showing the evolution of pneumonia during the therapy.

that of US, without a statistically significant dif- tions (fluid bronchograms or vessels) and by hy-
ference. perechoic linear structures with comet tails
In particular, two cases of small bronchopneu- (airbronchogram). Fluidbronchogram can be also
monia diagnosed with US, were considered with seen in obstructive pneumonitis (Yang et al.,
CR as interstitial pneumonia. Two other cases 1990), while partial airbronchogram is typical of
missed with CR, probably because small lesions pneumonia (Yang et al., 1990; Gehemacher et al.,
located in the left paracardiac region, were re- 1995).
vealed with US as small hypoechoic pneumonia. Sometimes this consolidation may be described
As regards the sonographic imaging of lobar or as a 'liver-like' US image, especially in the initial
sublobar pneumonia, it is characterized by a phase of the illness. Features of small bronchop-
wedge-shaped hypoechoic area with fine echoes neumonia show small subpleural hypoechoic le-
inside, interrupted by anechoic canalicular forma- sions, roundish or triangular-shaped, with fine
echoes inside and delimited by the hyper-
Table 3 echogenous of the surrounded pulmonary par-
Radiological and US diagnosis in the two groups and diagnos-
tic sensitivity (DS) compared with the final diagnosis enchima.
In the detection of pleural effusions US is supe-
Method NOR IP (DS) AP (DS) PE (DS) rior to CR (Kelbel et al., 1991) and in our study
the difference between the two techniques seems
CR 22 25 (92%) 33 (90%0 3 (71%)
US 43 O 37 (100%) 5 (100%)
to be more significant, to the advantage of US,
Final diagnosis 20 23 37 5 but the numbers considered are too small to draw
definitive conclusions. On the other hand, US is
 A. Benci et al. / European Journal of Ultrasound 4 (1996) 169-176 175

not able to identify interstitial pneumonias, ment as a routine diagnostic method, as it is easy
whereas CR takes on an important role, as confi- to perform on bed-ridden patients and cheaper
rmed by the data of our study (diagnostic sensitiv- than CT. In immunocompromised patients in par-
ity of CR at 92%). ticular, US of the chest could play an important
This study shows that a US examination also is role in difficult cases and very ill patients.
useful in the therapeutic follow-up of pneumonias Finally, US can be employed in the echo-guided
and bronchopneumonias; indeed, it detects extraction of infectious focus material, thus per-
residues of illness even when radiology shows mitting a rapid and certain aetiologic diagnosis
pulmonary lesions to be healed. This was shown (Diefenthal et al., 1988).
to be the case in two of the 37 patients affected by
alveolar forms (5.4%).
Sonomorphological differential diagnosis of References
pneumonia should include abscess, pulmonary in-
farction, atelectasis and lung tumors. Peripheral Diefenthal HC, Tashijian J. The role of the plain films, CT,
tomography, ultrasound and percutaneous needle aspira-
abscess often show characteristic features, such as
tion in the diagnosis of inflammatory lung disease. Semin
irregular borders and inhomogeneous hypoechoic Respir Infect 1988; 3: 83-105.
pattern, due to air artefacts or, rarely, hydro- Foy HM, Cooney MK, Allan I e t al. Rates of pneumonia
pneumonia level. Pulmonary infarction has the during influenza epidemics in Seattle, 1964 to 1975. J Am
same US pattern of pneumonia, that is, triangu- Med Assoc 1979; 241: 253-258.
lar, well-defined and hypoechoic lesions, but does Gehemacher O, Mathis G, Kopf A et al. Ultrasound imaging
of pneumonia. Ultrasound Med Biol 1995; 21: I 119-1122.
not show airbronchogram. Moreover, color Gehemacher O, Mathis G. Colour doppler sonography of
Doppler may be useful for differential diagnosis peripheral pulmonary consolidation. Imaging 1994; 61:1 I.
(Gehemacher and Mathis, 1994; Yuan et al., Hsu WH, Chang CD, Hsu JY et al. Detection of mass lesion
1994). Atelectasis, due to voluminous pleural effu- in the collapsed lung by ultrasonography. Taiwan I Hsueh
sion, shows pneumonic consolidation, which is Hui Ysa Chih 1992; 91: 57-62.
Kelbel C, Bomer N, Schadmand Set al. Diagnostica of pleural
often biconcave, floating in the pleural flid. Fi- effusion and atelectases: a comparison of sonography and
nally, lung tumors may present different sono- radiology. Frtschr Rontgenstr 1991; 154: 159-163.
graphic patterns: roundish, policyclic shape, Kroschel U, Seitz K, Reuss J, et al. Sonographic imaging of
regular or spicular borders, and homogeneous or lung emboli. Results of a prospective study. Ultraschall
Med 1991; 12: 263-268.
inhomogeneous texture; lung tumors, causing ob-
Jokinen C, Hieskanen L, Juvonen H, et al. Incidence of
struction, appear either an hypoechoic nodule community acquired pneumonia in the population of four
near the hylum or as a hyperechoic mass inside minicipalities in Eastern Finland. Am J Epidemiol 1993;
the consolidated lung (Yang et al., 1990). 5:269-268.
US, therefore, can play a significant role in the Mandell LA, Community acquired pneumonia. Chest t995;
diagnosis of infectious pathologies, which affect 108: 355-423.
Meduri GU, Stein DS. Pulmonary manifestations of Acquired
the pulmonary alveolus and can also be rationally immunodeficiency syndrome. Clin Infect Dis 1992; 14: 98.
employed in the diagnosis and in the monitoring Mathis G, Metzler J, Fussengger G, et al. Ultrasound findings
of the therapy of pneumonia and bronchial pneu- in pneumonia. Ultraschall Klin Prax. 1992; I l: 667-672.
monia during pregnancy and in infancy, thus Mathis G, Difschmid K. Pulmonary infarction: sonographic
appearance with pathologic correlation. Eur J Radiol 1993;
reducing the risks caused by repeated radio-
17: 170-174.
graphic examinations. Mikloweit P, Zachgo W, Lorcher U et al. Para-pleural lung
Moreover, US seems to be more sensitive than lesions: diagnostic value of sonographic versus computer-
CR in the diagnosis of minor pleural effusions ized tomography. Bildgebung 1991; 58: 127-131.
and to be equally, if not more, sensitive in the Miller JA, Lee S, Lao R et al. Comparison of thallium-201
and gallium-67 citrate scintigraphy in the diagnosis of
diagnosis of alveolar lesions, encouraging its em-
pulmonary disease. Chest 1995; 107: 440-446.
ployment if not routinely at least in doubtful Primack SL, Muller NL. High resolution computed tomogra-
cases, which appear CR negative. In Italy US is phy in acute diffuse lung disease in the immunocompro-
widely available and used by nearly every depart- mised patient. Radiol Clin North Am 1994; 32:731-744.
176 A. Benci et al. / European Journal o f Ultrasound 4 (1996) 169-176

Targhetta R, Burgeois JM, Marty-Double C et al. Towards Yang PC, Chang DB, Yu C J, et al. Ultrasound percuta-
another diagnostic approach of peripherial pulmonary neous cutting biopsy for the diagnosis of pulmonary con-
masses in pneumology: ultrasonically guided needle punc- solidations of unknown aetiology. Thorax 1992; 47:
ture. Rev Pneumol Clin 1991; 47: 2-8. 457 460.
Wassie E, Buscombe JR, Miller RF et al. 67Ga scintigraphy in Yang PC, Luh KT, Wu HD, et al. Lung tumors associated
HIV positive patients: a review of its clinical usefulness. Br with obstructive pneumonitis: US studies. Radiology
J Radiol 1994; 67:349-352. 1990; 174: 717-720.
Wu HD, Yang PC, Lee LN. Differentiation of lung abscess Yuan A, Chang DB, Yu CJ, et al. Colour doppler sonogra-
and empyema by ultrasonography. Taiwan I Hsueh Hui phy of benign and malignant pulmonary masses. A JR
Tsa Chih 1991; 90:749 754. 1994; 163: 545-549.