Use of lung ultrasound in connective tissue disease associated interstitial lung disease

Abstract

Pulmonary involvement is a common manifestation of various rheumatologic diseases, significantly impacting prognosis. Early diagnosis is crucial, and the recent use of bedside ultrasound by clinicians has opened new diagnostic avenues. This review explores the current state of the art regarding rheumatologic diseases and their ultrasound representation.

Resumen

El compromiso pulmonar es una manifestación frecuente de diferentes enfermedades reumatológicas y genera un impacto importante en el pronóstico. Su diagnóstico temprano es fundamental y el uso reciente de la ecografía a la cabecera del paciente por parte del clínico representa un nuevo campo diagnóstico. En esta revisión se plantea el estado del arte actual de la enfermedad reumatológica y su representación ecográfica.

Introduction

Over time, lung ultrasound has emerged as a valuable tool in assessing pulmonary pathology, including pleural effusion. It is currently integrated into clinical practice and recognized within international guidelines, such as the 2010 recommendations of the British Thoracic Society (BTS) [1]. Its portability and the continuous improvement in image quality have facilitated its integration into the diagnostic approach to connective tissue disease associated interstitial lung disease (CTD-ILD), serving as a strategy to guide therapeutic interventions and as a follow-up tool [2]. Its advantages and limitations are summarized in Table 1. This review discusses recent advancements in this area.

Table 1.

Advantages and disadvantages of lung ultrasound.

Advantages	Disadvantages
Wide availability and portability of the equipment	Operator dependent
Low cost	Not feasible in patients with subcutaneous emphysema
Absence of ionizing radiation	Limited applicability in patients with narrow intercostal spaces
Real time guidance for procedures

Adapted from Hassan et al. [2].

Technical aspects of ultrasound

Ultrasound transducers emit sound waves with a specific frequency and wavelength. These waves penetrate tissues, and a portion of them are reflected to the transducer, where they are detected via the piezoelectric effect, generating a two-dimensional image. This image is based on the time it takes for the waves to return, thereby indicating the distance of the structure from the surface. During this interaction, some energy is lost, and not all waves return to the transducer – a phenomenon known as attenuation. Certain structures with high attenuation coefficients, such as bone and air, significantly impact image quality and interpretation [1,2].

Ultrasound frequency is inversely related to depth and directly correlated with image resolution. Various transducers with different frequency ranges and designs are available, allowing for the assessment of several anatomical structures [1,2]. For the assessment of the chest wall and parietal pleura, high-frequency linear transducers (7.5–30 MHz) are useful, providing detailed imaging at depths of 2–5 cm. In contrast, the evaluation of deeper structures, such as pleural effusions and pulmonary abnormalities, requires low-frequency transducers (2–5 MHz), typically convex, with a penetration depth of 10–25 cm from the skin [1,2]. However, as will be discussed later in the context of interstitial lung disease assessment, studies have been conducted using both types of transducers, demonstrating a high degree of similarity. This finding supports the use of either transducer in these clinical scenarios [3].

Image acquisition

The transducer should be positioned between the ribs, perpendicular to the intercostal space, with the probe marker oriented toward the patient's head. In this ultrasound window, a hyperechoic structure appears in the center of the screen, corresponding to the pleural line, which exhibits movement known as pleural sliding (Video 1). At the edges of the field, the rib cage and its corresponding acoustic shadow can be observed, forming what is known as the “bat sign.”

In normal conditions, multiple hyperechoic lines can be seen repeating below the pleural line at regular intervals. These represent a reverberation artifact resulting from the interaction of ultrasound waves with air, specifically within the alveoli. These artifacts are referred to as A-lines [1].

Ultrasound assessment of the interstitial syndrome

International evidence-based guidelines on lung ultrasound, first established in 2012 and reaffirmed in the updated 2023 guidelines, emphasize the need to categorize positive ultrasound findings. This has led to the definition of interstitial syndrome, which refers to the occupation of the interstitial space and the loss of the normal alveolar air pattern, without specifying the nature of the occupying material (e.g., fluid such as blood, transudate, inflammatory exudate, or cellular infiltration) [4,5].

To identify interstitial syndrome, it is essential to understand the following definitions:

B-lines: These are hyperechoic reverberation artifacts that appear vertically on the ultrasound screen. They originate from the pleura and extend to the bottom of the screen. In the case of interstitial syndrome, these B-lines are believed to result from the interaction of sound waves with the interlobular septa (Fig. 1) (Video 1) [4,5].

Fig. 1.

Lung ultrasound using a linear transducer. Red arrow: pleural irregularity. Blue arrow: A-line. Green arrow: B-line.

Pleural abnormality: Defined as the evidence of fragmentation, irregularity, or thickening of the continuous hyperechoic line, an artifact generated by the friction between the visceral and parietal pleura during respiratory movements (Fig. 1) (Video 1) [4,5].

These findings align with those agreed upon in the OMERACT ultrasound working group's standardization for interstitial lung disease (ILD), with a consensus agreement level of 82.6% for pleural line irregularity and 84.2% for B-lines. The inter-observer reliability was assessed with a kappa correlation coefficient of 0.51 (0.39–0.64) for B-lines and 0.58 (0.43–0.74) for pleural line irregularity, while intra-observer reliability showed a kappa of 0.72 (0.67–0.78) for B-lines and 0.75 (0.69–0.81) for pleural line irregularity [6].

The 2012 international consensus establishes that the signs suggesting an interstitial syndrome are [4]:

• •

  The presence of three or more B-lines in the longitudinal plane between two ribs.

• •

  Two or more bilaterally positive regions.

In the 2023 guidelines, in addition to the aforementioned findings, it is suggested to standardize the description of these based on the extent of lung involvement, whether focal or diffuse, concerning its appearance, whether homogeneous or heterogeneous, and whether there is an apical-caudal or ventral–dorsal concentration gradient. Regarding the pleural line, it is recommended to include characteristics such as irregularity, fragmentation, or thickening [5].

Based on the findings, probable diagnoses can be proposed. When diffuse bilateral involvement is observed, it may suggest causes such as pulmonary edema, diffuse inflammatory pulmonary involvement (toxicity, infection), or diffuse ILD. In contrast, focal involvement of B-lines may be present in a normal lung or in conditions such as pneumonia, atelectasis, contusion, pulmonary infarction, pleural disease, or neoplasia [4].

For evaluation, the ideal approach involves the examination of eight anterolateral regions (four per hemithorax). However, two other possible methods are suggested: the assessment of 28 intercostal spaces, as described by Jambrick, using a systematic approach where intercostal spaces are assessed on the anterior chest on both sides beneath the parasternal, midclavicular, anterior axillary, and midaxillary lines, with some spaces on the left side being unassessable due to the interference of the heart (Jambrick); or the evaluation of two anterior regions in each hemithorax if time availability is limited [7].

However, specifically for the identification of ILD, protocols have evolved, starting with the expansion of the evaluation to a 50-point protocol, which included posterior points as described by Gargani et al. [8], it has been used and validated in the studies by Tardella et al. and Barskova et al. [9,10]; However, a significant drawback is the need to explore so many areas, requiring more time, which is why Gutiérrez et al. [11] developed a shortened protocol, where intercostal spaces were used in relation to the division of the thorax with vertical lines according to anatomical landmarks in the anterior, lateral, and posterior regions as follows: second parasternal space, fourth midclavicular space, fourth anterior axillary space, fourth midaxillary space, eighth paravertebral space, eighth subscapular space, and eighth posterior axillary space on both hemithoraces, totaling 14 locations (Fig. 2); a comparison of this protocol was made with that of Gargani et al. [8] and with the semi-quantitative index for assessing the extent and severity of fibrosis in the Warrick tomography (CT), showing an inter-observer kappa coefficient of 0.76–0.88 and an intra-observer kappa of 0.85–0.95, with an average examination duration of 8 min, allowing the conclusion that the 14-point protocol is a viable alternative [11].

Fig. 2.

Simplified 14-point protocol for the assessment of interstitial lung disease in patients with autoimmunity. Examined points: 2nd intercostal space (ICS) – parasternal, 4th ICS – midclavicular (MC), 4th ICS – anterior axillary (AA), 4th ICS – midaxillary (AM), 8th ICS – posterior axillary (AP), 8th ICS – subscapular, 8th ICS – paravertebral. Each point is assessed bilaterally. Abbreviations: MC, midclavicular line; AP, posterior axillary line; AM, midaxillary line; AA, anterior axillary line; ICS, intercostal space.

Adapted from: Gutiérrez et al. [11]. Image created with: BioRender.

Ultrasound findings related to the severity of CTD-ILD

In 2016, Buda et al. proposed a more objective semi-quantitative system for ultrasound findings, where B-line numbers are scored as <3, ≤4, and “white lung” (>10 B-lines), which correlate with ground-glass opacity on CT. The irregular pleural line is also assessed, with a new descriptor – “blurred” or poorly defined – associated with greater extension. Additionally, new lines termed “Am” are identified, linked to subpleural cysts, with the lesion score being the sum of the number of affected segments [12].

In a study of 84 patients with ILD and idiopathic pulmonary fibrosis, associated with autoimmune diseases and exposure-related pneumopathies, lung ultrasound was performed and compared with CT. The findings that helped differentiate the severity of fibrotic involvement, especially in the severe fibrosis group, included bilateral diffuse pleural thickening >3 mm (100%), sub pleural cysts (100%), decreased pleural sliding (61%) and increased B-lines (76%). In contrast, the mild involvement group only showed pleural thickening at the lung base, without any of the other mentioned signs [13].

Currently, high-resolution chest tomography is considered the non-invasive diagnostic method of choice for ILD, with a high correlation to histology. For this reason, lung ultrasound has been compared to this procedure.

In a study of 114 patients who attended a radiology department for high-resolution chest tomography without prior knowledge of any specific pulmonary or cardiovascular pathology, ILD was diagnosed in 39 (34%) of the cases. Lung ultrasound was performed on the same patients, without knowledge of the tomography diagnosis, finding B-lines in 91 cases (82%). The sensitivity was calculated at 89.7% (75.8–97.1%) and specificity at 70.7% (59–80%), with a negative predictive value of 93% for the identification of ILD [14].

In 66 patients suspected of having ILD, a comparison was made between ultrasound and high-resolution chest tomography. It was found that for the finding of ≥5 B-lines present in two of the examined areas, ultrasound had a sensitivity of 93% and specificity of 73%, compared to tomography, which had a sensitivity of 100% and specificity of 82%. The area under the curve showed no difference in the diagnostic value of the two methods for ILD (p = 0.48) [15].

Additionally, the utility of ultrasound in diagnosing CTD-ILD has been studied. In 40 patients with systemic sclerosis (SSc), it was found that the presence of 10 or more B-lines correlated with the presence of significant ILD on high-resolution tomography, with a Warrick index ≥7, a positive likelihood ratio (LR+) of 12.52, sensitivity of 96%, and specificity of 92%. As an additional result, it was found that the presence of 5 or more B-lines, while having high sensitivity (100%), showed low specificity (69%) for ILD, prompting a reconsideration of the significant number for SSc-ILD [16].

In some groups, the number of B-lines has been correlated with other clinical findings, especially in the population with systemic sclerosis, such as a decrease in DLCO of 73.6 ± 16.5 (p < 0.0001) or the presence of digital ulcers, with an average of 29 ± 21.8 B-lines (p < 0.01) [17].

For the reasons mentioned above, we consider that lung ultrasound has become a useful, practical, and valid tool for the study of ILD at the bedside.

Lung ultrasound (LUS) as a screening strategy for ILD in patients with SSc

ILD is one of the leading causes of morbidity and mortality SSc, and its early diagnosis is essential for impacting disease progression, with high-resolution chest computed tomography (HRCT) currently being the gold standard for diagnosis [18]. However, due to the ionizing radiation involved and the potential limitations in access to this technique, in recent years, the utility of LUS for screening ILD in SSc has been evaluated. It allows for assessment during the consultation, free from radiation and at a lower cost [19]. Considering the OMERACT (Outcome Measures in Rheumatology) filters, the evidence in the literature supports its apparent and content validity, as well as its feasibility. However, other aspects such as criterion validity, reliability, and sensitivity to change still require further investigation [20].

The ability of LUS to detect the presence of ILD and quantify its extent has been analyzed through the assessment of B-lines and/or alterations in the pleural line [21]. The majority of studies have analyzed B-lines and have shown a strong correlation with the presence and severity of ILD on HRCT [22–33]. Although the data predominantly comes from patients with advanced SSc, similar findings have been described in early stages of SSc and even in patients without respiratory symptoms or abnormalities in pulmonary function tests [22,23,29].

The utility of pleural line abnormalities for screening ILD in SSc has been less extensively studied, but preliminary data are of significant interest. These data demonstrate a strong correlation with B-line indices, as well as with the presence, extent, and severity of ILD on HRCT [34]. It has been reported that a pleural line thickness greater than 3 mm can identify subclinical ILD in patients with SSc who show no abnormalities in pulmonary function tests or the 6-minute walk test (6MWT). Furthermore, the presence of subpleural nodules and the degree of pleural line thickening have been suggested to correlate with different ILD patterns on HRCT [30–32].

Some authors propose that pleural line abnormalities have greater accuracy than B-lines in identifying the presence of ILD [30–32] whether measuring pleural line thickening or using qualitative indices, such as that of Fairchild et al. [35] or semi-quantitative [31,32,36]. Additionally, while B-lines can be identified in approximately 7% of healthy controls, pleural line abnormalities are typically absent [30].

Good intra- and interobserver agreement has been reported for the study of B-lines and pleural line abnormalities, regardless of the type of probe (sector, convex, or linear) and ultrasound machine (high-end or handheld) used. However, recent data suggest better interobserver reliability for pleural line abnormalities compared to B-lines, with Cohen's kappa values of 0.84 and 0.78, respectively [36].

In patients with SSc, B-lines have been inversely associated with forced vital capacity (FVC), total lung capacity (TLC), and carbon monoxide diffusion capacity (DLCO), and directly associated with clinical and laboratory parameters linked to disease severity or the risk of developing ILD, such as anti-topoisomerase-1 (Scl70) antibodies, diffuse skin involvement, progression of capillary damage on capillaroscopy, digital ulcers, and severity index [22,25–27,29,31]. A significant correlation has also been described between B-lines and KL-6 (Krebs von den Lungen-6), a recognized biomarker of ILD [37]. Pleural line abnormalities have also shown an inverse correlation with FVC and TLC, as well as a significant association with Scl70 antibodies [36].

Supporting the potential utility of LUS for screening SSc-ILD are its high sensitivity (60–100%) and negative predictive value (NPV) (52–100%). Additionally, B-lines demonstrate notable specificity (34–89%) and positive predictive value (PPV) (78–95%). Similarly, pleural line abnormalities also exhibit high sensitivity (74–100%) and specificity (82–99%) [21]. The breadth of these ranges may be related to the considerable heterogeneity among studies, not only in the characteristics of the included populations but also in ultrasound examination protocols, probe types, ultrasound indices, and the cutoff values used.

The superior performance of LUS in terms of sensitivity, specificity, PPV, and NPV (91.2%, 88.6%, 87.0%, and 92.4%, respectively) is noteworthy when compared to commonly used clinical tools such as chest radiography (2.5%, 98.1%, 40.0%, and 66.6%, respectively), pulmonary function tests (27.5%, 77.3%, 41.4%, and 64.7%, respectively), or the presence of velcro crackles on auscultation (8.7%, 98.3%, 41.6%, and 87.5%, respectively) [29].

The various proposed indices for B-lines and pleural line abnormalities – whether qualitative, semi-quantitative, or quantitative – incorporate different intercostal space (ICS) counts with their respective cutoff values [34]. All these aspects still require consensus to ensure the optimal implementation of LUS in clinical practice for ILD screening in patients with SSc. Further validation as an outcome measure is necessary.

LUS as a screening tool for ILD in patients with rheumatoid arthritis (RA)

The existing evidence in the literature regarding the utility of lung ultrasound as a screening tool for ILD in patients with RA is significantly more limited than for SSc, and its validation is less advanced. However, the findings remain equally promising [21]. Unlike in SSc, ILD-RA is less prevalent and less incident, typically manifesting at later stages of the disease, although it can also present as an initial manifestation in some patients. Despite advances in understanding the risk factors associated with ILD development in RA, there is still no universally accepted consensus on its screening. In this context, having a non-invasive tool such as lung ultrasound could be highly beneficial.

As in SSc, most studies have focused on the assessment of B-lines using convex, sectorial, or linear probes, with fewer data available on pleural line abnormalities. While early studies had limited sample sizes (n = 39–75), more recent data include larger and more informative cohorts (n = 155–196) [39–48]. All data refer to patients with advanced rheumatoid arthritis, with a mean disease duration ranging from 8 to 14 years.

A significantly higher presence of B lines has been demonstrated in patients with RA compared to healthy controls (28% vs. 7%, respectively) [40]. Similarly to what has been described for SSc, pleural line abnormalities may also be more precise in discriminating ILD, as they were present in the form of fragmentation, nodules, or pleural thickening in 4%, 18%, and 28% of patients with RA, respectively, but were not identified in healthy controls [40]. Furthermore, a significantly higher number of B lines has been demonstrated in patients with RA and associated ILD compared to patients with RA without ILD involvement [43,44].

Some studies have investigated the utility of lung ultrasound as an additional screening tool in patients with suspected ILD involvement based on symptoms, signs, or radiological abnormalities, while other studies have also included asymptomatic patients (34–56.6%) recruited consecutively [44–49]. In all of these studies, B lines and/or pleural line abnormalities have shown a strong correlation with findings from high-resolution CT, demonstrating good accuracy: sensitivity (62.2–100%), specificity (55.6–100%), PPV (29.4–94.3%), NPV (69.5–98.6%), and area under the curve (AUC) (0.82–0.91) [40–49].

Due to the good accuracy demonstrated by reduced intercostal spaces (ICSs) counts in SSc, most studies have examined the 14 ICSs described by Gutiérrez et al. [38], although two manuscripts have investigated including the 72 EICs described by Gargani et al. [22] which has allowed us to obtain information of interest. Mena-Vázquez et al. [43] demonstrated that the number of B lines is significantly associated with the presence of ILD on HRCT, regardless of the number of ICSs examined, although the identified cutoff points for each ICS count were different. When exploring 72 ICSs, the presence of at least 10 B lines discriminated ILD with a sensitivity of 91.4% and a NPV of 86.9%. In contrast, when evaluating 10 ICSs, the cutoff was set at least 5.5 B lines, with acceptable sensitivity (68.6%) and NPV (69.5%). These authors proposed an even more reduced count by selecting the ICSs independently associated with the presence of ILD, aiming to shorten the ultrasound examination time: 2nd left medioclavicular ICS, 3rd right anterior axillary ICS, 8th right subscapular ICS, and 9th right paravertebral ICS (cutoff of 5.5 B lines; sensitivity 62.2% and NPV 69.5%). Additionally, Cogliati et al. [41], examining 50 ICSs, they demonstrated a good correlation (kappa 0.78) between the examination conducted by an expert sonographer using a conventional ultrasound machine and that performed by a less experienced examiner using a handheld ultrasound device (sensitivity 92% vs. 89% and specificity 56% vs. 50%, respectively).

In RA, B lines have also shown a significant correlation with the presence of antibodies (rheumatoid factor (RF) and anti-citrullinated cyclic peptides (aCCP)), disease activity as measured by DAS28, functional capacity as reflected by HAQ, KL-6, age, or male sex, and a negative association with FVC, FEV1, or DLCO [42–47,50]. Pleural line abnormalities have also been significantly correlated with age, RF, aCCP, IL-6, and male sex.

Of great interest are the comparative analyses of the accuracy of ultrasound (sensitivity 90.6–98.3%; NPV 84.6–94.7%) for detecting ILD versus other techniques such as plain radiography (sensitivity 29.9–64.5%; NPV 46.4%) or FVC, DLCO, and velcro crackles (sensitivity 28.1%, 63.3%, and 68.8%, respectively). These findings provide additional arguments in favor of using LUS for ILD screening in RA [48,49].

Use of LUS in diagnosis and screening of other autoimmune diseases

Although much of the research to date has focused on the use of ultrasound in autoimmune diseases such as RA and SSc, there are other autoimmune conditions in which this technique could play a significant role in identifying pulmonary involvement [51–53].

Autoimmune diseases such as anti-synthetase syndrome (ASyS), dermatomyositis (DM), and juvenile idiopathic arthritis (JIA) are characterized by potential pulmonary involvement, which may include interstitial pulmonary fibrosis and decreased respiratory function [53–56]. Up to 80% of patients with inflammatory myopathies (IIM) may be asymptomatic and present with diffuse ILD, making early diagnosis and continuous monitoring of these conditions crucial for improving patient prognosis and quality of life [54]. Additionally, these patients require regular evaluations for monitoring pulmonary complications, and lung ultrasound provides a safe and cost-effective alternative to radiation-based techniques, such as HRCT [57,58].

Several studies have demonstrated a significant correlation between ultrasound findings, such as the presence of B lines and pleural abnormalities, with results obtained from HRCT and other pulmonary function tests [59,60]. Furthermore, highlighting its high sensitivity for identifying early abnormalities, enabling clinicians to act more quickly and efficiently by adjusting treatment based on non-invasive, real-time findings [51,52].

A literature search was conducted to identify clinical studies published until September 2024 that evaluated the use of LUS in autoimmune diseases, defining its applications, benefits, and challenges in clinical practice. Eleven publications on this topic were identified (see Supplementary Material).

Systemic juvenile idiopathic arthritis (sJIA)

This technique has been explored for the detection of ILD in pediatric patients with sJIA. In a study conducted by Vega-Fernández et al. [56], feasibility of LUSs for screening pulmonary disease associated with sJIA was evaluated in comparison with HRCT. This technique revealed pleural and subpleural abnormalities that correlated with findings on HRCT, including pleural irregularities and scattered B lines [56]. This study highlights the utility of ultrasound in a pediatric context, where minimizing radiation exposure is particularly important.

Systemic lupus erythematosus (SLE)

In patients with SLE, ultrasound has proven useful in detecting abnormalities such as pleural effusion and thickening, which may be indicative of lupus pneumonitis or pleuritis [53]. Ultrasound demonstrates high sensitivity compared to chest X-rays for detecting these findings, being a non-invasive, radiation-free technique that allows for rapid assessment [53,55]. Additionally, the ease of access and the ability to detect abnormalities in real-time make ultrasound a valuable tool in the ongoing management of patients with SLE, where continuous monitoring of pulmonary complications is essential [53,55].

ANCA-associated vasculitis (AAVs)

Both, granulomatosis with polyangiitis (GPA) and microscopic polyangiitis (MPA), involvement of the lower respiratory tract can be observed [58]. A study compared LUS and HRCT findings in patients with these conditions, identifying lesions such as nodules, infiltrates, alveolar hemorrhage, and pulmonary fibrosis in 33 of 35 patients. The results from both methods were consistent, with only four showing variability. However, this underscores the need for further studies with a larger number of patients [58].

Polymyositis and dermatomyositis

In polymyositis (PM) and DM, pulmonary manifestations include ILD [54,55]. Some studies have explored the use of lung ultrasound for the early detection of ILD, showing that it can identify B lines (artifacts associated with the presence of fluid in the interstitial space) [59–61]. Although HRCT remains the gold standard for diagnosing ILD, ultrasound has proven to be a useful screening method in patients with limitations for undergoing a HRCT scan [53].

Additionally, Mongodi et al. demonstrated that lung ultrasound can be used as a complementary method to HRCT for monitoring acute exacerbations of ILD involvement associated with DM–PM, guiding immunosuppressive therapy and evaluating treatment response continuously, without the need for frequent radiation exposure [62]. This combined approach is particularly advantageous for patients who require close monitoring without the risk associated with cumulative radiation exposure.

Furthermore, a significant correlation has been found between the use of B lines in lung ultrasound and serum levels of KL-6 as markers of the severity of ILD in patients with IIM, suggesting these two markers as useful tools for monitoring the progression of ILD associated with IIM [60]. Similarly, in ASyS, a strong correlation has been found between B lines on lung ultrasound and ground-glass opacities on HRCT [59].

Additionally, a case report described a patient with anti-MDA5 positive DM who developed respiratory failure due to the combination of ILD and diaphragmatic myositis [57]. In this case, lung and diaphragmatic ultrasound were used to assess respiratory dysfunction, revealing significant diaphragmatic hypomobility [57].

In another case report, a patient with clinically amyopathic DM associated with anti-MDA-5 positive-ILD was monitored using LUS and the serum marker KL-6 [61]. The study concluded that the combination of LUS with serum biomarkers such as KL-6 provides a non-invasive and effective tool for monitoring these patients, allowing for early assessment of pulmonary changes and treatment response [61].

Sjögren's syndrome (SjS)

A pilot study evaluated the accuracy of LUS for diagnosing ILD in patients with SjS, correlating the results with HRCT, which is considered the non-invasive gold standard for diagnosing ILD. It was found that LUS has a sensitivity of 1 (95% CI 0.398–1), specificity of 0.89 (95% CI 0.518–0.997), and a positive likelihood ratio of 9.00 (95% CI 7.1–11.3) for detecting pulmonary pathology. These results suggest an excellent correlation with HRCT in patients with SjS affected by ILD and could be a useful technique in daily clinical practice for evaluating pulmonary disease in dry syndrome [63].

Our analysis of the studies found that most used a prospective observational design or systematic review, providing a robust level of evidence for the use of LUS in autoimmune diseases. Additionally, it was observed that the use of LUS can be particularly beneficial in clinical settings where access to more advanced techniques, such as HRCT, is limited [53–56,63]. Beyond its utility in detecting pulmonary disease, LUS stood out for its ability to be performed frequently, making it ideal for long-term monitoring in children [53,56]. The main limitation of LUS in these conditions is its operator dependency and the lack of standardization in diagnostic criteria [51]. In this regard, the operator's experience plays a critical role in obtaining and interpreting the images, which may lead to variability in the results [51]. Moreover, most of the available studies are small-scale and focus on specific cases, which makes it difficult to generalize the results [56–58,60,62].

LUS as a follow-up or response to treatment in autoimmune-associated ILD

The information regarding the follow-up of autoimmune diseases with interstitial involvement is limited.

A literature search was conducted to identify clinical studies published up until September 2024 that evaluated the use of LUS in the follow-up or treatment of ILD associated with autoimmunity, and 4 publications addressing this specific question were identified. The LUS findings from the studies retrieved through this search are described below, with representative figures provided.

Currently, under the recommendations of the American College of Rheumatology (ACR), follow-up with HRCT is proposed based on clinical condition and risk factors [64]. To date, the role of LUS in monitoring the condition and assessing therapy response has not been clearly defined; however, there is some preliminary data on this.

The group led by Gutiérrez et al., conducted a validation study using LUS to determine the utility of this technique in the progression of ILD associated with ILD-SSc. They compared 133 patients without respiratory symptoms to 133 healthy controls, evaluating Borg scores, Rodnan scores, auscultation, chest radiography, and pulmonary function tests. Continuous follow-ups were performed every 12 weeks up to 48 weeks. Fifty-nine point four of the patients showed abnormal ultrasound findings compared to 4% of healthy controls, which was statistically significant (p = 0.0001), indicating a high prevalence of the disease despite being asymptomatic. The abnormal ultrasound findings demonstrated a good correlation with HRCT (rho = 0.802; p = 0.0001) and a concordance rate of 90.2% between the two evaluation methods, documenting a LUS sensitivity of 91.2% and specificity of 88.6%. Six cases were identified as false positives and seven as false negatives. At the one-year follow-up, only 121 patients were evaluated. Of the 79 patients with abnormal LUS findings, 30 (37.97%) demonstrated progression of ILD associated with SSc at 12 months, as assessed by semi-quantitative lung ultrasound evaluation. Of these, only 30% developed symptoms during the follow-up, and 40–55% showed deterioration in pulmonary function tests (excluding the diffusion capacity of the lungs for carbon monoxide, DLCO). However, the authors note that it was not possible to perform HRCT on all patients. Given these data, ultrasound is proposed as an important strategy for monitoring patients in preclinical or subclinical stages as a risk predictor for detecting disease progression. However, external validation is still needed, and it remains to be determined whether this finding is only applicable to ILD-SSc or if it behaves similarly in other conditions associated with ILD, such as IIM, RA, or SjS [65].

The research group led by Battista et al. conducted an exploratory study on LUS changes in relation to nintedanib (NIN) therapy strategies in SSc with pulmonary involvement [66], they conducted imaging follow-up with HRCT at baseline and at 12 months, using LUS every 3 months for one year, along with clinical follow-up, functional studies including DLCO and spirometry. Their objective was to evaluate the quantitative changes in LUS and HRCT in patients with SSc-ILD. As a secondary aim, they sought to assess any correlation between imaging changes and the efficacy of NIN, based on functional outcomes and quality of life. The study involved 10 patients. LUS was performed by an expert with 5 years of experience, blinded to the HRCT findings, in 55 areas, with the presence of B-lines or pleural thickening considered abnormal. Regarding the patients, 70% were women, with a mean age of 62 years, and 30% were active smokers. At the beginning of the follow-up, 100% were on a combined immunosuppressive therapy plus NIN, and by the end of the follow-up, only 40% continued with this regimen. The remaining 60% continued treatment with NIN. One patient (10%) discontinued therapy due to severe gastrointestinal intolerance, although 66% experienced diarrhea, which improved with dose adjustments. Among the patients who continued the therapy, all showed stabilization of FVC. Interestingly, a reduction in B-line count was documented, from 175.1 (66.7) to 120.8 (70.3), which was statistically significant (p = 0.005). Similarly, pleural irregularities decreased over the follow-up period, from 50.6 (32.5) at baseline to 37.2 (22.4) (p = 0.05). These findings suggest that lung ultrasound could serve as a potential follow-up strategy to assess the efficacy of the antifibrotic therapy NIN in patients with SSc-ILD [66].

Regarding therapy response, in 2022, the group of D’Orazio et al., emphasized the need for strategies to determine whether LUS could be used as a monitoring tool in SSc patients who are refractory to first-line management. To this end, the researchers studied changes in pleural thickening and its modification in response to therapy with or without biologic disease-modifying antirheumatic drugs (bDMARDs) or with the use of mycophenolate mofetil (MPM) and rituximab (RTX). They conducted a prospective follow-up of 29 patients, who were simultaneously assessed with both HRCT and LUS. Their results were compared to a previous study involving 25 SSc patients, who were followed with HRCT and LUS but did not receive bDMARDs, serving as the control group. Among the 29 patients, 17 had evidence of SSc-ILD, and all showed a good response to treatment compared to those who did not receive the therapy. Pleural thickening was significantly lower in the group treated with bDMARDs (1.8 ± 0.7 mm vs. 0.95 ± 0.31 mm; p < 0.000). Similarly, these patients experienced fewer dry symptoms, less dyspnea, and a better prognosis. Among the group that did not initially receive bDMARDs but later started treatment, 5 patients showed a reduction in pleural thickening (mean 0.68 ± 0.2 vs. 1.5 ± 0.4; p < 0.000). The authors concluded that these findings suggest that LUS could be used as a follow-up strategy to assess treatment response in patients with SSc-ILD [67].

Current evidence suggests that this emerging technology in the rheumatologist's diagnostic arsenal may be a useful tool for monitoring patients with rheumatologic diseases and associated pulmonary involvement. However, the studies remain exploratory and do not represent some of the key groups in rheumatologic disease (e.g., IIM, SjS, and SLE) that could be of interest in the future. Additional studies are needed to determine its role not only as a screening strategy but also as a clinical follow-up tool.

Conclusion

LUS has become a highly useful and accessible tool for screening, diagnosis, monitoring, and prognosis of ILD involvement in autoimmune diseases. The growing and robust evidence supporting its use in the rheumatology outpatient setting correlates with clinical and paraclinical markers of disease activity across various autoimmune conditions. It evaluates several ultrasound parameters and features; however, the 14 intercostal space protocol, focusing on the identification of B-lines and pleural irregularities, remains the most sensitive and specific finding for diagnosing and assessing the severity of interstitial pulmonary involvement. Although the majority of evidence comes from SSc, studies in RA, SLE, SjS, IIM, AAV and sJIA also support its use, owing to its low cost, excellent performance, and absence of radiation exposure. Today's rheumatologist must be sufficiently trained to incorporate this tool into their daily practice.

CRediT authorship contribution statement

All authors contributed to the conception and design of the study. An initial meeting, organized by areas of focus, was held to structure the topics to be addressed. Each author contributed equally to the drafting of the manuscript.

Ethical considerations

This study was conducted as a bibliographic review and did not involve patient data; therefore, informed consent was not required.

Funding

The authors received no financial support for this study.

Conflict of interest

The authors declare no conflicts of interest.