Imaging findings of pulmonary contusions on multidetector CT: A retrospective study comparing adults and children

Abstract

To describe imaging findings of pulmonary contusions (PC) in adults and children using multidetector computed tomography (CT) scanners. We conducted a retrospective single center study. All chest multidetector computed tomography (MDCT) scans of victims of blunt trauma admitted to the emergency unit of a reference trauma center of Brazil between January 2015 and December 2016 were reviewed in search of opacities compatible with PC. The CT images were analyzed in conjunction with medical records, that provided demographic and clinical data. The obtained data were analyzed in the overall population and comparing children and adults. Significant P value was defined as <.05. 52.7% of patients presented bilateral opacities. Middle third, posterior and peripheral portions of the lungs were more frequently affected, in the craniocaudal, anteroposterior and axial axes, respectively. A vast majority of patients (80.6%) presented multiple opacities, whereas a minority showed subpleural sparing (26.9%) and fissure crossing (22.6%), with similar frequencies in children and adults. Children, although, more frequently presented consolidation and more diffuse lesions in the anteroposterior axis compared to adults, with statistically significant differences. PC usually are multiple and predominate in middle, posterior and peripheral portions of the lungs. Subpleural sparing and fissure crossing seems to be infrequent and have similar frequencies between children and adults. Although, there are differences between these age groups, as younger people tend to have more consolidation and diffuse opacities in the anteroposterior axis than older ones.

1. Introduction

Pulmonary contusions (PC) are parenchymal lesions caused by rupture of capillaries in the alveolar and septal walls with blood leakage into alveolar space and into interstitium.^[1] They constitute the most common pulmonary lesion in blunt chest trauma, occurring in 17% to 75% of patients,^[2] also being important cause of morbidity and mortality, this one estimated at 10% to 25%, depending on severity and presence of associated lesions.^[3] Besides that, they are an independent risk factor for development of acute respiratory distress syndrome and pneumonia.^[4]

The first description of PC related to blunt trauma has been attributed to Morgagni in 1761^[5] and, since then, they are topic of medical interest. The first descriptions of radiographic findings date back to the first half of the last century.^[6] Currently, the methods most used for the diagnosis are radiography and computed tomography (CT), this with a higher sensitivity,^[7–10] especially in the initial 6 hours after trauma.^[11] An experimental study conducted in dogs demonstrated that immediately after induced trauma radiographs have a sensitivity of 38%, while CT has a sensitivity of 100% to detect PC.^[12] Therefore, CT can be assumed as a gold standard method.^[13–16] The image pattern described in the literature is patchy or nonsegmental alveolar opacities, ranging from patchy ill-defined ground-glass opacities (GGO) to extensive areas of mixed lesions or consolidations.^[2] Among the characteristics that can help to differentiate it from others airspace diseases are subpleural sparing, which is described as the noninvolvement of 1 to 2 mm of pulmonary parenchyma beneath the pleura^[17] and their tendency to not respect anatomic limits, including pulmonary fissures, since they originate from the application of external forces on the pulmonary parenchyma.^[18] Despite all previous knowledge, there are few data on frequency of imaging findings and some statements about this issue were grounded only from authors’ observation and experience. Besides that, the vast majority studies were performed before widespread use of multidetector computed tomography (MDCT) scanners.^[19] Here, we describe the findings of MDCT in PC and associated lesions, comparing the frequency of findings in children and adults.

2. Materials and Methods

2.1. Ethics statement

This study was approved by the local Human Research Ethics Committee and informed consent was waived.

2.2. Design and setting

We conducted a retrospective single center study. All chest MDCT scans of victims of blunt trauma admitted to the emergency unit of a reference trauma center of Brazil between January 2015 and December 2016 were reviewed in search of opacities compatible with PC. The CT images were analyzed in conjunction with medical records, that provided demographic and clinical data.

2.3. Participants

Initially 94 patients were selected, but one with incomplete medical records was excluded, lasting a total of 93 subjects. 40 of them presented pulmonary lacerations along with isolated PC.

2.4. Imaging technique

All patients were examined with a 16-channel GE Lightspeed MDCT scanner (General Electric, Milwaukee, WI, USA) and the images were stored in Picture Archiving and Communication System, where they were obtained from for analysis by radiologists at workstations. Typical imaging parameters were: 120 to 140 kVp; 100 to 300 mAs; 1.25 mm slice thickness; and pitch equals 0.9375 to 1.375.

2.5. Imaging interpretation

After initial training, the images were separately evaluated by 2 radiologists with 5 and 7 years of experience in chest radiology, respectively. Independent tabulated reports were generated and the cases in which there was disagreement between the observers were again analyzed by a third radiologist with 35 years of experience in chest radiology in order to reach consensus. All opacities adjacent to pulmonary lacerations were not considered.

Lesions distribution was assessed in 3 axes: craniocaudal, anteroposterior and axial. For evaluation of the first we have considered the upper, middle and lower thirds of the lungs; for evaluation of the second we have divided the lungs into anterior and posterior halves and classified the findings as anterior and predominantly anterior, posterior and predominantly posterior, and without predominance; and for assessment of axial distribution we have divided the lesions into central and peripheral (the last involving the external third of the lungs and the others considered as central) and classified the findings into central and predominantly central, peripheral and predominantly peripheral and without predominance. Opacities types were classified as GGO, consolidation or mixed lesions (GGO associated to consolidation). Subpleural sparing was considered when up to 2 mm of the subpleural parenchyma was normal adjacent to peripheral opacities. Fissure crossing was considered when there were subpleural lesions in adjacent areas of different lobes of the same lung with similar morphology to each other.

2.6. Statistical analysis

All statistical analyses were performed using the computer program Stata/ SE v.14.1. StataCorpLP, USA. Categorical variables were described by frequencies and percentages and continuous variables by mean and standard deviation (SD). We used Fisher exact test to compare categorical variables and independent t test to compare continuous variables. P value <.05 was considered statistically significant.

3. Results

3.1. Patient demographics and clinical data

Ninety-three participants were included, comprising 12 children with a mean age of 10.6 years old (range 2–17 years), and 81 adults, with a mean age of 34.1 years old (range 18–90 years). All patients underwent MDCT scans up to 6 hours and 34 minutes after admission to the emergency unit. The mean injury severity scale was 22.25 (SD = 12.60) for children and 22.20 (SD = 10.35) for adults, with no statistically significant difference between these groups.

Table 1 has a description of the study subjects with respect to sex and injury mechanism. The classification “others” includes assaults, bicycle falls, horse falls and impact of blunt objects. Traffic accidents (motorcycle fall, car crash and auto-pedestrian accident pooled) were the most frequent causes in children and adults (91.7% and 66.7%, respectively). Despite that, children were much more frequently victims of auto-pedestrian accidents, while adults suffered motorcycle accidents in a higher frequency. Also, there was predominance of falls in adults.

Table 1.

Number and percentages of patients according to sex and injury mechanism.

Variable	Classification	Total (n = 93)	Children (n = 12)	Adults (n = 81)
Sex	Female	13 (14.0%)	3 (25.0%)	10 (12.3%)
	Male	80 (86.0%)	9 (75.0%)	71 (87.7%)
Injury mechanism	Fall	18 (19.4%)	1 (8.3%)	17 (21.0%)
	Motorcycle accident	28 (30.1%)	1 (8.3%)	27 (33.3%)
	Car crash	22 (23.7%)	3 (25.0%)	19 (23.5%)
	Auto-pedestrian accident	15 (16.1%)	7 (58.3%)	8 (9.9%)
	Others	10 (10.7%)	0 (0.0%)	10 (12.3%)

3.2. Distribution of opacities

There was a slight predominance of bilateral lesions overall. However, motorcycle accidents presented more frequently with unilateral lesions (57.1%) while car crash, auto-pedestrian accidents and falls presented higher percentages of bilateral ones (54.5%, 60.0%, and 61.1%, respectively).

In the craniocaudal distribution all thirds were affected in 45.2% of the patients and in the rest of them the middle third was the most frequently involved, alone or in conjunction with the other thirds. In its turn, the posterior half and the periphery were generally affected, when we considered the isolated or predominant involvement of these portions of the lungs. Children usually have more nonpredominant distribution of opacities in the anteroposterior direction when compared to adults, with statistically significant difference between these age groups (Table 2).

Table 2.

Laterality and distribution of opacities in craniocaudal, anteroposterior, and axial axes.

Variable	Classification	Total (n = 93)	Children (n = 12)	Adults (n = 81)	P
Laterality	Unilateral	44 (47.3%)	4 (33.3%)	40 (49.4%)	.364
	Bilateral	49 (52.7%)	8 (66.7%)	41 (50.6%)
Craniocaudal distribution	Superior third	8 (8.6%)	1 (8.3%)	7 (8.6%)
	Middle third	10 (10.7%)	2 (16.7%)	8 (9.9%)
	Inferior third	8 (8.6%)	0 (0.0%)	8 (9.9%)
	Superior and middle thirds	8 (8.6%)	1 (8.3%)	7 (8.6%)
	Middle and inferior thirds	16 (17.2%)	0 (0.0%)	16 (19.8%)
	Inferior and superior thirds	1 (1.1%)	0 (0.0%)	1 (1.2%)
	All thirds	42 (45.2%)	8 (66.7%)	34 (42.0%)
Craniocaudal distribution	Superior third affected	59 (63.4%)	10 (83.3%)	49 (60.5%)	.199
	Middle third affected	76 (81.7%)	11 (91.7%)	65 (80.2%)	.688
	Inferior third affected	67 (72.0%)	8 (66.7%)	59 (72.8%)	.733
	All thirds affected	42 (45.2%)	8 (66.7%)	34 (42.0%)	.130
Predominant anterior/posterior distribution	Anterior	21 (22.6%)	0 (0.0%)	21 (25.9%)
	Posterior	49 (52.7%)	4 (33.3%)	45 (55.5%)
	Without predominance	23 (24.7%)	8 (66.7%)	15 (18.5%)
	With predominance	70 (75.3%)	4 (33.3)	66 (81.5%)	.001
	Without predominance	23 (24.7%)	8 (66.7%)	15 (18.5%)
Predominant axial distribution	Central	12 (12.9%)	3 (25.0%)	9 (11.1%)
	Peripheral	50 (53.8%)	6 (50.0%)	44 (54.3%)
	Without predominance	31 (33.3%)	3 (25.0%)	28 (34.6%)
	With predominance	62 (66.7%)	9 (75%)	53 (65.4%)	.745
	Without prominence	31 (33.3%)	3 (25%)	28 (34.6%)

3.3. Number of opacities

The vast majority (80.6%) of patients had more than one lesion. Isolated lesions were uncommon (Table 3).

Table 3.

Amount and types of opacities, subpleural sparing and fissure crossing.

Variable	Total (n = 93)	Children (n = 12)	Adults (n = 81)	P
Only one opacity	18 (19.4%)	3 (25.0%)	15 (18.5%)	.696
Only GGO	44 (47.3%)	1 (8.3%)	43 (53.1%)	.004
Subpleural sparing	25 (26.9%)	4 (33.3%)	21 (25.9%)	.728
Fissure crossing	21 (22.6%)	2 (16.7%)	19 (19.2%)	.728

GGO = ground-glass opacities.

3.4. Types of opacities

47.3% of patients had only GGO, while 52.7% had at least one mixed opacity or homogeneous consolidation. Comparing children and adults showed statistically significant difference between these groups: adults presented much more only GGO opacities than children (Table 3).

3.5. Subpleural sparing and fissure crossing

These 2 characteristics were present in the minority of patients (26.9% had subpleural sparing and 22.6% had fissure crossing and), as is shown in Table 3. Comparing children and adults for both evidenced no statistically significant differences of their frequencies.

3.6. Associated lesions

Rib fractures, pneumothorax, pulmonary laceration and pleural effusion were the more incident associated chest lesions. They were more frequent in adults, nevertheless without statistically significant difference. Other types of lesions were much less frequent and are summarized in Table 4. 16 (36.4%) from 44 patients with unilateral PC had also unilateral rib fractures, 14 ipsilateral and 2 contralateral.

Table 4.

Associated lesions.

Variable	Total (n = 93)	Children (n = 12)	Adults (n = 81)	P
Atelectasis	9 (9.7%)	1 (8.3%)	8 (9.9%)	1.000
Pulmonary laceration	40 (43.1%)	6 (50.0%)	34 (42.0%)	.756
Pneumothorax	41 (44.1%)	4 (33.3%)	37 (45.7%)	.540
Pleural effusion	14 (15.1%)	1 (8.3%)	13 (16.0%)	.685
Pneumomediastinum	9 (9.7%)	1 (8.3%)	8 (9.9%)	1.000
Rib fracture	44 (44.1%)	5 (41.7%)	39 (48.1%)	.763
Scapular fracture	12 (12.9%)	0 (0.0%)	12 (14.8%)	.353
Clavicle fracture	11 (11.8%)	2 (16.7%)	9 (11.1%)	.630
Sternal fracture	2 (2.2%)	0 (0.0%)	2 (2.5%)	1.000
Dorsal spine fracture	13 (14.0%)	1 (8.3%)	12 (14.8%)	1.000
Airway lesion	2 (2.2%)	0 (0.0%)	2 (2.5%)	1.000
Flail chest	2 (2.2%)	0 (0.0%)	2 (2.5%)	1.000
Pneumomediastinum	9 (9.7%)	1 (8.3%)	8 (9.9%)	1.000
Extrathoracic lesions	78 (83.9%)	10 (83.3%)	68 (84.0%)	1.000

4. Discussion

4.1. Patient demographics and clinical data

In scientific literature, the main mechanism of injury in PC is traffic accident,^[20,21] whilst children are more frequently victims of auto-pedestrian accidents and adults are prone to suffer more car crashes,^[22–27] such as we have evidenced here.

4.2. Distribution of opacities

Most of PC were bilateral, what is in disagreement with previous studies, that evidence a slight^[16,28] or vast^[23,29] predominance of unilateral PC. Bilateral PC are more frequent in children who suffer traffic accidents compared to others injury mechanisms^[26] and our data showed that falls generally were related to bilateral lesions and responsible for a significant percentage of PC. Indeed, different percentages of injury mechanisms in our population may help to explain different numbers of unilateral and bilateral PC compared to other studies.

Opacities were present in all thirds of the lungs in almost half of the cases, but when they predominate in one or 2 thirds, the middle third was more frequently involved. This is in disagreement with what other authors, given that they have assumed that PC are usually located in the inferior third.^[30,31] Although an increased basilar mobility is supposed to be an important cause of lung laceration,^[32] the obtained data may show that this mechanism could be less important generating isolated PC.^[33]

The most of subjects had posterior or predominantly posterior located lesions. A previous study^[17] has previously showed that lesions are predominantly posterior in children and our data agreed with these results in both age groups. There are 2 possible explanations to the posterior predominance: the most flexible anterior wall of the chest associated to contrecoup forces may compress the posterior portion of the lungs against the firm posterior wall and spine; and the curved posterior chest wall may act as a mirror after an anterior impact, reflecting and concentrating the force on the posterior portions of the lungs (Fig. 1).^[22] Nevertheless, comparing predominant and nonpredominant lesions, in the anteroposterior direction children had more lesions without predominance than adults, with statistically significant difference. This may be related to a trend of younger patients to have more diffuse lesions, at least in the referred axis. This propensity may be associated to children’s smaller body, that makes the forces to be distributed over less mass.^[22,27]

Figure 1.

Crescentic GGO in the subpleural posterior portion of the inferior right lung of a 26 years old man victim of a motorcycle accident. This type of opacity is possibly associated to reflection of forces by the chest wall (“mirror effect”). GGO = ground-glass opacities.

The propensity of PC to be peripheral (usually near solid structures such vertebrae, ribs, liver and heart), as we have noticed here, was observed by a lot of authors,^[6,33–41] albeit they did not have evaluated the frequency of this finding. That occurrence can be explained by the transmission of forces through the chest wall reaching mainly the peripheral parts of lungs.^[2] Also, the mentioned mechanism of chest wall acting as a mirror and causing reflection of forces to the adjacent portions of the lungs may help to explain this peripheral predominance, considering that it can occur in any part of the thoracic circumference.

4.3. Number of opacities

We have verified that the great majority of patients have multiple lesions. There is no mention to this variable in the previous scientific literature, at least from what we know.

4.3.1. Types of opacities.

The frequencies of types of opacities were not evaluated in previous studies, as far as we know, and he have noticed here that there are similar frequencies of only GGO and mixed lesion/consolidation in adults, with slight predominance of the first group, while children usually have at least one mixed lesion or consolidation. As younger people have less calcified bones and more flexible ligaments, they have a more pliable chest wall,^[14,20,27] which allows extensive bowing without fracture, directly transferring energy to the underlying lung with much less protection conferred by the chest wall.^[5,42] It may explain why they have more mixed lesions or consolidations, given that these opacities are possibly related to a more severe damage.^[2,43]

4.3.2. Subpleural sparing and fissure crossing.

Subpleural sparing (Fig. 2) was found in 95% of children in a previous study,^[17] a very higher percentage compared to what was obtained here. Fissure crossing (Fig. 3) was also infrequent, but we do not know of any study that have analyzed it before. These 2 imaging findings enable reliable differentiation of PC from others causes of lung opacification, as pneumonia and atelectasis,^[35] but they were infrequent, what makes them less useful in daily practice.

Figure 2.

Subpleural sparing (arrows) in 3 different patients: (A) 20 years old woman victim of car crash; (B) 13 years old male children victim of auto-pedestrian accident; and (C) 27 years old woman victim of run auto-pedestrian accident. There are motion artifacts in (B).

Figure 3.

Fissure crossing in a 27 years old man victim of motorcycle accident. (a) Axial and (b) sagittal planes. The arrows show the oblique fissure.

4.4. Associated lesions

Other thoracic extrapulmonary injuries were usually more frequent in adults (except for clavicle fractures). Although there were no statistically significant differences between the age groups, it may reveal a trend. These data agree with previous literature and is probably related to the referred characteristics of chest wall in children, that allow damages in lungs without significative lesions of thoracic wall.^{[14,22–24,40,42,44–47]} Besides that, greenstick fractures, typical of children, are also hard to identify and can be overlooked^[25]; however, MDCT tend to minimize this mistake.

Presence of rib fractures in contralateral hemithorax, as was showed in 2 cases, is curious, but may be explained by a contrecoup mechanism.^[5,35,48]

5. Limitations

The retrospective nature of the study and the relatively small number of pediatric patients are limitations. The nonspecific nature of lung opacity in a patient with trauma must also be recognized as a problem. To minimize that, medical records were carefully revised to exclude patients with clinical suspicion of preexisting pneumonia or fever at presentation. Besides that, all examinations were performed early, what makes PC the most probable cause of opacity. Finally, compressed files stored in Digital Imaging and Communications in Medicine server were susceptible to loss of quality, what may have affected analysis.

6. Conclusions

PC usually are multiple and predominate in posterior and peripheral portions of lungs, but they are usually present in all thirds in the craniocaudal axis or, when this is not the case, tend to involve the middle third. Subpleural sparing and fissure crossing are infrequent and have similar frequencies between children and adults. Although, compared to adults, children tend to have more frequently consolidation and distribution of opacities without predominance in the anteroposterior axis.

Author contributions

Conceptualization: Camilo Dallagnol, Juan Marcelo Fernandez Alcala, Dante Luiz Escuissato.

Data curation: Camilo Dallagnol, Juan Marcelo Fernandez Alcala, Rafaelle Machado de Vargas, Dante Luiz Escuissato.

Formal analysis: Camilo Dallagnol, Rafaelle Machado de Vargas, Dante Luiz Escuissato.

Funding acquisition: Camilo Dallagnol.

Investigation: Camilo Dallagnol, Juan Marcelo Fernandez Alcala, Dante Luiz Escuissato.

Methodology: Camilo Dallagnol, Juan Marcelo Fernandez Alcala.

Project administration: Camilo Dallagnol, Rafaelle Machado de Vargas.

Resources: Camilo Dallagnol.

Supervision: Dante Luiz Escuissato.

Validation: Camilo Dallagnol, Dante Luiz Escuissato.

Visualization: Camilo Dallagnol, Dante Luiz Escuissato.

Writing – original draft: Camilo Dallagnol.

Writing – review & editing: Camilo Dallagnol, Juan Marcelo Fernandez Alcala.