Transthoracic ultrasonographic features of diffuse parenchymal lung diseases

Objectives The value of transthoracic ultrasonography (TTUS) in diffuse parenchymal lung disease (DPLD) has not been established yet. This prospective study was conducted to assess pleural and parenchymal alterations in patients with DPLD by TTUS and to compare the results with chest high-resolution computed tomography (HRCT). The results of TTUS were compared with some arterial blood gas (ABG) parameters and 6-min walk test (6MWT). Patients and methods This study was conducted on 50 patients. All patients underwent HRCT, resting ABGs, and 6MWT, in addition to TTUS to evaluate (a) the presence and number of B-line and distance between them, (b) pleural effusion, (c) pleural thickening, (d) an irregular, fragmented pleural line, and (e) subpleural alterations. Results The studied group showed female predominance, wide range of age, and most of them were nonsmokers. All patients had diffuse bilateral B-lines. B-line numbers were inversely correlated with severity of pulmonary fibrosis detected by HRCT modified Warrick score; however, distance between B-lines was directly correlated with severity of pulmonary fibrosis. Some ABGs (resting PaO2, resting SpO2%, AaDO2) and 6MWT parameters were correlated with B-line number and distance between them. As detected by TTUS, the majority of patients (82%) had irregular thickened pleural line, whereas 44% of them had absent lung sliding. Conclusion TTUS can play a complementary role in the diagnosis and monitoring of DPLD patients. Multiple B-lines distributed over the entire lung surface in combination with a thickened, irregular, and fragmented pleural line are strongly suggestive of the presence of DPLD. TTUS gives an idea about diffuse or limited, early, or advanced DPLD.


Introduction
In recent years, transthoracic lung ultrasound (TTUS) has risen as a new, promising technique for assessment of numerous pulmonary abnormalities [1]. It has been highlighted that lung sonography is highly sensitive to variations of the pulmonary content and balance between air and fluids. Of late, lung US has been considered a valid method for the assessment of interstitial pulmonary fibrosis compared with high-resolution computed tomography (HRCT) [2]. Interstitial lung disease (ILD) can be imaged by the presence of B-lines (also called US lung comets). B-lines are a sonographic sign of interstitial syndrome, originating from thickened interlobular septa [3].

Aim
The aim of this study was to assess pleural and parenchymal alterations in patients with diffuse parenchymal lung disease (DPLD) by TTUS and to compare the results with the severity of changes in chest HRCT. In addition, the results of TTUS examination were compared with some arterial blood gas (ABG) parameters and 6-min walk test (6MWT), which reflect the degree of pulmonary impairment secondary to fibrosis.

Patients and methods
This descriptive cross-sectional study was conducted at our institute. The ethical committee of university hospital institute approved the study. An informed written consent was obtained from all participants before their enrolment into the study. Fifty patients with DPLD were included in this study. They were selected from the chest department and the chest outpatient clinic. All patients were diagnosed according to diagnostic algorithm of ILD [4]. The diagnosis of these cases was based on clinical presentation, serological markers, pulmonary function tests, HRCT pattern and distribution, bronchoscopy with bronchoalveolar lavage, and biopsy [Transthoracic Ultrasonography (TTUS) or by mediastinoscopy].

Exclusion criteria
Patients with pulmonary edema of various causes, pneumonia, atelectasis, chronic obstructive pulmonary disease, pulmonary embolism, and pleural disease were excluded from the study, as B-lines may be detected in all mentioned cases [3,5,6] Study design All patients were subjected to the following: (1) Detailed medical history and clinical examination, including demographics, family, smoking, drug, environmental, and occupational exposures history, with emphasis on the onset, course, and duration of the presenting symptoms, and the symptoms and signs of extrapulmonary systems affection [4]. (2) Laboratory tests: these tests included complete blood count, erythrocyte sedimentation rate, urine analysis, fasting and postprandial blood sugar, liver function tests, creatinine, blood urea nitrogen and serological markers such as antinuclear antibody, lupus erythematosus cells, rheumatoid factor, cytoplasmic antineutrophil cytoplasmic antibodies, perinuclear antineutrophil cytoplasmic antibodies, and antiscleroderma 70 antibodies (when indicated according to clinical diagnosis of the case) [4]. (3) Resting ABGs analysis: ABGs were analyzed by 'rapid lap analyzer 248' apparatus; the following values were recorded and then calculated: O 2 saturation% (SaO 2 %), PaO 2 mmHg, PaCO 2 mmHg, HCO 3 mEq/l and pH, and AaO 2 gradient. (4) Six-min walk distance (6MWD): it is supervised measurement of the distance that patients can quickly walk on a flat, hard surface in a period of 6 min. 6MWD was carried out before and after interventions to determine the patient functional capacity [7]. Reference values are 576 m for healthy male patients and 494 m for healthy female patients [8]. Before and after the test, oxygen saturation% (SpO 2 %) was measured by a pulse oximeter, and desaturation was defined as a fall in oxygen saturation to 88% or less during the 6MWT [9]. (5) HRCT HRCT examination was performed by standard protocol using Aquilion 160 helical scanner (Toshiba, Otawara, Japan). Scans were obtained at full inspiration from the apex to the lung base with the patients in the supine position. The acquisition parameters were as follows: sequential mode, 1-1.5 mm collimation and 10 mm interval, 180-260 mA average tube current (depending on body built), and 120-140 kV tube voltage. In the current study, pulmonary involvement was identified and scored according to a semiquantitative modified Warrick score [10]. Modified Warrick scoring was obtained by summing the point values of parenchymal abnormality pattern (severity score) ranging from 0 (normal) to 15 (all lesions present) and (the extension score) by localizing the number of pulmonary segments involved of both the lungs and adding +1 for each segment involved. An additional score of +3 was assigned when the lesions were bilateral. The extension score ranged from 0 (normal) to 15 (more than nine pulmonary segments involved). The total Warrick score ranged from 0 (no involvement) to 30 (the worst involvement). The results were expressed as normal (0 point); 1, mild (>8 points); 2, moderate (8-15 points); and 3, severe (<15-30 points) ( China SONOMED) equipped with a 3.5-MHz convex probe and 8-MHz linear probe. The patient was examined in a supine or near supine position with the anterior chest wall exposed rising his/her arms above his/her head to widen the intercostal spaces and facilitate scanning. We used convex probe and then linear probe in the same manner for effective assessment of the pleura; scanning was performed using the intercostal spaces as acoustic windows.
In the evaluation of interstitial fibrosis, thoracic sonography was performed for each patient using the protocol described by Volpicelli et al. [3] and Anderson et al. [2]. The sonographic technique ideally consists of scanning eight zones. Four chest areas per side, areas 1 and 2 on right side, as well as areas 5 and 6 on left side denote the upper anterior and lower anterior chest areas, respectively. Areas 3 and 4 on the right side, as well as areas 7 and 8 on left side, denote the upper lateral and basal lateral chest areas, respectively ( Fig. 1).

TTUS interpretation:
(1) B-lines (a) The presence of B-lines, which were previously called comet-tail artifacts, generated from the thickened interlobular septa at the lung wall interface was assessed. These were defined as well-defined, vertical, hyperechoic, dynamic lines originating from the pleural line, and spreading like a laser ray up to the edge of the screen. These artifacts are best visible under real-time examination, because some of them appear less pronounced on frozen sonograms. A positive region is defined by the presence of three or more B-lines in a longitudinal plane between two ribs and a positive exam is defined by the presence of two or more positive regions bilaterally [3]. The numbers of B-lines per scan were counted and then the total number of B-lines in all previously mentioned areas was calculated [3].The distance between each two adjacent B-lines was measured and then the average was calculated and was expressed in millimeters. (c) Detection of presence or absence of pleural nodule or pleural effusion. (d) Estimation of pleural thickness: we measured the thickness of pleural lines in all zones and then calculated the mean value. Pleura was considered thickened if its thickness was more than 3 mm [11].

Statistical analysis
Data were analyzed by the statistical package for the social science software, version 20, Armonk, NY: IBM Corp. Descriptive results were expressed as mean±SD, and non-numerical data were expressed as number and percentage. Graphics were performed using Microsoft excel 2010. χ 2 analysis was used to compare between TTUS and HRCT data. Study of the relationship between variables was performed using 'Pearson correlation'. Linear correlation coefficient (r) was used for the detection of the correlation between two quantitative variables in one group. Both the direction and strength of relation were expressed by calculating the correlation coefficient (r). The strength is expressed by a number ranging between 0 (absent correlation) and 1 (perfect correlation). In practice, it is usually fractions where the more close to 1 the stronger the correlation. The following can serve as a guide to interpret the magnitude of correlation coefficient.

Results
Fifty patients with DPLD were included in this study (34 women and 16 men with extended age ranging from 26 to 83 years, mean age of 51.92±13.28 years); 19 patients were recently diagnosed with DPLD (38%, n=19), and 31 patients were previously diagnosed with DPLD and were attending the outpatient clinic for follow-up (62%, n=31). Clinical characteristics and demographic data of the study group are summarized in Table 2. This table shows that large portions of DPLD participant patients (80%) were nonsmokers, and presented with more than one symptom. All of them (100%) presented with dyspnea and cough, followed by chest wheeze in 58% of patients. Additional extrapulmonary manifestations were recorded in 22% of patients. The majority of our patients were diagnosed with hypersensitivity pneumonitis (HP) (54%, n=27), and sarcoidosis was recorded in 8% (n=4).
To assess severity of pulmonary fibrosis, the patients were divided into three groups (mild, moderate, and severe) according to HRCT semiquantitative modified Warrick score (Table 3) [10]. Most of our patients were classified as moderate (48%, n=24), followed by severe (34%, n=17), and lastly mild (18%, n=9).  Table 4 shows total B-line number, mean B-distance, and pleural thickness for all studied patients as mean± SD. In addition, this table shows that most of the cases had thickened and irregular pleural line (62 and 82%, respectively), whereas 44% had reduced or absent lung sliding. Table 5 and Fig. 3 show that the degree of interstitial affection on chest HRCT detected by modified Warrick score is significantly inversely correlated with total B-line number and positively correlated with B-line distance between each two adjacent Blines among all patients involved-mild, moderate, and severe groups. Table 6 and Fig. 4 show that there was no statistical significant difference in total number of B-lines and Bline distance detected by TTUS between mild, moderate, and severe total Warrick score (P=0.79 and 0.73, respectively), whereas there was a significant difference in pleural thickness of all patients involved in each group and pleural thickness of patients with thickened pleura in each group (P=0.012 and .026, respectively). Table 7 shows that there was no statistical significant difference in resting PaO 2 , resting SaO 2 %, AaDO 2 , 6MWD, and post-6MWT SpO 2 % among all patients in mild, moderate, and severe groups. Table 8 shows that B-line number is inversely correlated with B-line distance in mm and AaO 2 gradient, whereas it is positively correlated with resting PaO 2 , resting SpO 2 %, post-6MWT SpO 2 %, and 6MWD among all patients involved − mild, moderate, and severe groups. Table 9 shows that B-line distance is positively correlated with AaO 2 gradient, whereas it is negatively correlated with resting PaO 2 , resting SpO 2 %, post-6MWT SpO 2 %, and 6MWD among all patients involved − mild, moderate, and severe groups (Fig. 2).

Discussion
Lung ultrasound is a reliable tool to evaluate DPLD. The presence of multiple B-lines is the primary sonographic sign in DPLD. Pleural line abnormalities are often present in patients with DPLD. The distribution of B-lines correlates with computed tomography signs of fibrosis [2].
In the current study, about 80% of DPLD participant patients were nonsmokers, and presented with more        patients were diagnosed with HP (54%, n=27), and sarcoidosis was recorded in 8% (n=4).
These results were in agreement with Behr [12], who concluded that dyspnea at rest or on exertion is a common presenting symptom in 90% of cases of ILD. Dry or productive cough of nonpurulent sputum is also a common feature in up to 80% of ILD patients. In addition, Leslie [13], reported that HP and sarcoidosis are less likely to occur in smokers.
In the current study, newly diagnosed cases of DPLD were 38% (n=19) and previously diagnosed DPLD attending outpatient clinic for follow up were 62% (n=31). All patients were diagnosed according to diagnostic algorithm of ILD [4]. Final diagnosis was either by clinical diagnosis, serological markers, pulmonary function tests, and HRCT pattern and distribution or by clinical diagnosis, HRCT, bronchoalveolar lavage, and biopsy [lymphnode biopsy through bronchoscopy (TBLB), or mediastinoscopy]. These results coincide with Launay et al. [14], who reported that for the diagnosis of ILD as it is a sensitive method to assess the extent and the pattern of pulmonary fibrosis. However, Kanne [15], concluded that a complete lack of pulmonary   Pleural assessment by TTUS in three studied groups. parenchymal changes on HRCT imaging virtually excludes a diagnosis of ILD. Moreover, Meyer [16], detailed that HRCT of the thorax could give data that strongly support a specific diagnosis and may be diagnostic [e.g. typical changes of usual interstitial pneumonia (UIP)] such that further testing with bronchoscopy or surgical lung biopsy is not required.
In patients with DPLD, the subpleural interlobular septa are thickened by deposition of collagen and fibrous tissues. By investigating the lung surface by TTUS, the great impedance gradient between the thickened septa and air in the lung causes reflection of ultrasound beams, creating diffuse B-lines all over the lung surface.
In our study, by TTUS examination, all patients had diffuse and more than three B-lines on both sides of the lung. These B-lines were best visible under real-time examination, as some of them seem less declared on frozen sonograms. Similarly, we observed that the Blines appear clearly visible and widely separated from each other (7 mm) corresponding to well-established pulmonary fibrosis with reticular pattern or honeycombing in HRCT, whereas there were numerous B-lines with a narrow distance between them (3 mm), making the lung more hyperechoic, giving the description of white lung that was observed corresponding to ground glass areas in HRCT.
These outcomes were in agreement with Hasan and Makhlouf [17], as they concluded that B-lines appear obvious and widely separated from each other in patients with well-established fibrosis in HRCT; in addition, the B-lines seem numerous with a narrow distance between them in those with early ILD. They proposed that numerous B-line artifacts with a narrow distance between them might be because of the fact that the area between two adjacent interlobular septa became filled with inflammatory cells and edema. Bouhemad et al. [21], also stated that multiple Blines 7 mm apart are caused by thickened interlobular septa, characterizing interstitial edema, whereas B-lines 3 mm or less apart are caused by ground glass areas characterizing alveolar edema. Same conclusions were also reached by Lichtenstein et al. [22].
In our study, the total B-lines number and mean Bdistance for all studied patients (mean±SD) were 72.1 ±9.05 and 4.12±0.76, respectively. Despite this, there was no significant difference in total number of B-lines and B-line distance identified by TTUS, between mild, moderate, and severe modified Warrick score (P=0.79 and 0.73 respectively); we found that the degree of interstitial affection on chest HRCT detected by modified Warrick score is significantly inversely correlated with total B-line number and positively correlated with B-line distance between each two adjacent B-lines among all patients involved − mild, moderate, and severe groups.
These results were in concurrence with Hasan and Makhlouf [17], as they reported that the distance between each two adjacent B-lines positively correlated with Warrick score (r=0.693) (P<0.001). Moreover, Cogliati et al. [18], concluded that the Bline score was significantly correlated with HRCT score (r=0.806). Nearly same results were obtained by Mohammadi et al. [19], as they reported that Warrick score was significantly correlated with severity of pulmonary involvement (r=0.695, P<0.001).
Different results were reported by Gargani et al. [20], who found that mean US B-lines for total studied patients with systemic sclerosis (SSc) was 37±50. This score is more frequent in the diffuse form of SSc rather than in the limited form (73±66 vs. 21±35; P<0.05) and has a good correlation with HRCT on assessment of lung fibrosis (r=0.72, P<0.001). Likewise, mean Warrick score was higher in the diffuse form than in the limited form (6.9±4.1 vs. 1.6±2.5; P<0.001). Therefore, they concluded that US B-lines have a potential diagnostic value to detect pulmonary fibrosis. The contrast between this outcome and our outcome is most presumably because they studied mean B-lines number and Warrick score in diffuse versus limited form of interstitial lung infiltration in SSc only, just while our study was on different types of DPLD.
In the present study, we used linear probe for more details about the pleural line. We recognized that pleural thickness in all studied patients as mean±SD was 3.205±0.34 mm. The majority of cases had thickened pleural line (62%, n=31), whereas 38% (n=19), had normal pleural thickness with mean thickness±SD as 3.36±0.35 and 2.95±0.05 mm, respectively. In addition, there was irregular and fragmented pleural line in most cases (82%, n=41), whereas 18% (n=9), of patients had regular pleural line. On the other hand, we identified that 44% (n=22), of our studied patients showed reduced or absent lung sliding, whereas 56% (n=28), had ordinary lung sliding.
Comparable results were accounted for by Reissig and Kroegel [23], as they found that 84.9% (n=45) of their DPLD cases had thickened pleural surface, whereas 98.1% (n=52) of cases had irregular pleural surface.
In the current study, despite the fact that there was no significant difference (P=0.09) in percentage of patients (mild, moderate, and severe groups) with thickened pleura distinguished by TTUS, there was a significant difference (P=0.012) in the pleural thickness in millimeter (mean±SD) between all groups involved. However, there was a significant difference in percentage of patients with irregular pleura and absent or reduced lung sliding between mild, moderate, and severe groups (P=0.002 and 0.002, respectively). Moreover, all severe cases showed irregular pleural line over the whole lung (100%), thickened pleura, and nearly absent lung sliding in the vast majority of them (83.3 and 70.6%, respectively). On the other hand, the moderate group showed less value for irregularity, thickening, and sliding of the pleura (83.3, 54.2, and 41.6%), and these abnormalities were identified essentially at the base of the lung. Conversely, all mild cases demonstrated normal lung sliding while they had irregular thickened pleural line just at the lung base in 44.4% of them. These outcomes showed that the pleural abnormalities were better distinguishable in the lower part of the lung, and the degree of fibrosis turned out to be more pronounced as the disease progress.
These results were in agreement with those of Sperandeo et al. [24]. They investigated the role of TTUS features of mild, moderate, and severe pulmonary fibrosis (based on clinical, PFT and radiological finding). They reported that severe pulmonary fibrosis cases showed irregular and thickened pleural line over the whole lung (100%, 33/33) with reduced lung sliding in 20/33 (61%). Moderate group showed 32/32 (100%) with irregular thickening at both lung bases with reduced lung sliding in 13/32 (41%). Mild cases showed normal lung sliding, whereas irregular pleural thickening was observed just at the lung base in 19/19 (100%). This slight contrast between Sperandeo et al. [24] study and our own might be because they had bigger studied cases (n=84 vs. 50) and majority of them were IPF. Furthermore, Soldati et al. [25], concluded that in DPLD the degree of the pleural involvement can vary within the same patient, and usually a thickened and fragmented pleural line with multiple B-line is better detectable in the lower part of the lung, which reflects the extent of the fibrosisIn the current study, there was no statistically significant difference in resting PaO 2 , resting SaO 2 %, AaDO 2 , 6MWD, and post-6MWT SpO 2 % among all studied groups. We have observed that all the parameters measured as mean±SD decrease as the disease has become more advanced, with the exception of increasing gradient (AaO 2 ) directly with the severity of the illness, which indicates more impairment in gas exchange.
In the current study, B-line number is inversely correlated with B-line distance in millimeter, and AaO 2 gradient, whereas it is positively correlated with resting PaO 2 , resting SpO 2 %, post-6MWT SpO 2 %, and 6MWD among all patients involved − mild, moderate, and severe groups. On the other hand, B-line distance is positively correlated with AaO 2 gradient, whereas it is negatively correlated with resting PaO 2 , resting SpO 2 %, post-6MWT SpO 2 %, and 6MWD among all patients involved − mild, moderate, and severe groups. These results indicated that PaO 2 , AaO 2 gradient, and 6MWD correlated with severity of DPLD, as numerous B-lines with narrow distances between them represent early alveolar wall affection and little impairment in lung function, whereas less B-line number with wide distances between them indicate thickened septa and great functional impairment.
These results were in agreement with Hasan and Makhlouf [17], as they reported that the distance between B-lines inversely correlated with PaO 2 (r=−1.811) (P<1.110).

Conclusion
Thoracic US is favored by lower cost, no radiation exposure and bedside examination. In addition, it can play a complementary role in the diagnosis and monitoring of DPLD patients, especially when HRCT unavailable or cannot be done. Diffuse bilateral B-lines distributed over the lung surface in combination with a thickened, irregular, and fragmented pleural line are strongly suggestive of the presence of DPLD. HRCT can give an idea about the probable histopathological types; however, TTUS gives an idea about diffuse or limited, early, or advanced DPLD.

Recommendations
Future or new blinded study to assess the sensitivity, specificity, and diagnostic accuracy of TTUS in patients with DPLD versus PCX-Ray and HRCT. Further studies are recommended to attain DPLD severity score assessment via TTUS and to compare them with other severity scores such as severity of symptoms, ABG, 6MWT, HRCT, spirometry, and DL CO .

Limitations
The study has a few limitations. First, TTUS had been act upon patients already diagnosed as DPLD, and this may be regarded as a bias for the interpretation. However, in this study, we do not assess the diagnostic accuracy of TTUS in DPLD patients, but we study the role of TTUS in evaluation of DPLD patients and whether it can play a complementary role in the diagnosis and follow-up of DPLD patients. Second, interstitial fibrosis is usually not homogeneously disseminated. This limitation is not considered, as the majority of the studied patients had diffuse disease and on TTUS examination nearly all of the affected areas were assessed.

Financial support and sponsorship
Nil.

Conflicts of interest
There are no conflicts of interest.