Imaging of Pulmonary Involvement in Rheumatic Disease




Lung disease commonly occurs in connective tissue diseases (CTD) and is an important cause of morbidity and mortality. Imaging is central to the evaluation of CTD-associated pulmonary complications. In this article, a general discussion of radiologic considerations is followed by a description of the pulmonary appearances in individual CTDs, and the imaging appearances of acute and nonacute pulmonary complications. The contribution of imaging to monitoring disease, evaluating treatment response, and prognostication is reviewed. Finally, we address the role of imaging in the challenging multidisciplinary evaluation of interstitial lung disease where there is an underlying suspicion of an undiagnosed CTD.


Key points








  • There is a high degree of overlap between the pulmonary manifestations of the various connective tissue diseases (CTDs), particularly with respect to interstitial lung disease (ILD) patterns.



  • Nonspecific interstitial pneumonia (NSIP) is the most common ILD pattern in all CTDs apart from rheumatoid arthritis, where usual interstitial pneumonia (UIP) is the most frequent pattern.



  • Distinguishing between acute exacerbation, infection, drug toxicity, or pulmonary hemorrhage as the cause for acute deterioration in CTD-associated lung disease is impossible on high-resolution computed tomography (HRCT) appearances alone, but requires the integration of clinical and serologic data with the evolution of appearances on plain radiograph.



  • A role for HRCT in staging disease and aiding prognostication has recently been shown, with traction bronchiectasis and extent of honeycombing both associated with increased mortality in CTD-ILD.






Introduction


The rheumatic diseases are a heterogeneous group of inflammatory disorders characterized principally by joint disease, but also, not infrequently, multiorgan dysfunction. Lung disease is common in connective tissue diseases (CTDs) and is an important cause of both morbidity and mortality. The CTDs (sometimes termed collagen vascular diseases) include rheumatoid arthritis (RA), systemic sclerosis (SSc), systemic lupus erythematosus (SLE), polymyositis/dermatomyositis (PM/DM), primary Sjögren syndrome (SS), mixed connective tissue disease (MCTD), and undifferentiated connective tissue disease (UCTD). Radiologic assessment has a definite role in the management of CTD-associated pulmonary disease. In this regard, plain chest radiography and high-resolution computed tomography (HRCT) are the principal tests. The indications for imaging will vary, but a typical scenario for the radiologist is to establish whether lung disease is present and, if so, to characterize its nature and extent. In cases in which a histospecific radiologic diagnosis of lung disease cannot be provided, HRCT may be the best guide to the optimal site for surgical biopsy. More recently, there has been interest in the utility of HRCT to assess longitudinal behavior and prognosis.


In the present article, a general discussion of radiologic features is followed by a description of the appearances in individual CTDs. We then consider the pulmonary complications of CTDs on imaging, the use of imaging in prognostication, and the role of multidisciplinary evaluation. Given the acknowledged limitations of plain chest radiography in characterizing patterns of diffuse lung disease in general and in CTD-associated lung disorders, the article focuses primarily on HRCT appearances. Pulmonary involvement also occurs in other rheumatic disorders, such as vasculitides and inflammatory disorders, including spondyloarthropathy, Behçet disease, and relapsing polychondritis, but the thoracic radiologic manifestations of these conditions are distinct from the CTDs listed previously and are outside the scope of this article.




Introduction


The rheumatic diseases are a heterogeneous group of inflammatory disorders characterized principally by joint disease, but also, not infrequently, multiorgan dysfunction. Lung disease is common in connective tissue diseases (CTDs) and is an important cause of both morbidity and mortality. The CTDs (sometimes termed collagen vascular diseases) include rheumatoid arthritis (RA), systemic sclerosis (SSc), systemic lupus erythematosus (SLE), polymyositis/dermatomyositis (PM/DM), primary Sjögren syndrome (SS), mixed connective tissue disease (MCTD), and undifferentiated connective tissue disease (UCTD). Radiologic assessment has a definite role in the management of CTD-associated pulmonary disease. In this regard, plain chest radiography and high-resolution computed tomography (HRCT) are the principal tests. The indications for imaging will vary, but a typical scenario for the radiologist is to establish whether lung disease is present and, if so, to characterize its nature and extent. In cases in which a histospecific radiologic diagnosis of lung disease cannot be provided, HRCT may be the best guide to the optimal site for surgical biopsy. More recently, there has been interest in the utility of HRCT to assess longitudinal behavior and prognosis.


In the present article, a general discussion of radiologic features is followed by a description of the appearances in individual CTDs. We then consider the pulmonary complications of CTDs on imaging, the use of imaging in prognostication, and the role of multidisciplinary evaluation. Given the acknowledged limitations of plain chest radiography in characterizing patterns of diffuse lung disease in general and in CTD-associated lung disorders, the article focuses primarily on HRCT appearances. Pulmonary involvement also occurs in other rheumatic disorders, such as vasculitides and inflammatory disorders, including spondyloarthropathy, Behçet disease, and relapsing polychondritis, but the thoracic radiologic manifestations of these conditions are distinct from the CTDs listed previously and are outside the scope of this article.




General radiologic and pathologic considerations


The CTDs can affect the pulmonary and extrapulmonary components of the thorax to varying degrees. The main manifestations in the pulmonary interstitium, airspaces, airways, pulmonary vasculature, pleura, pericardium, heart, mediastinum, and thoracic musculature are given in Table 1 . Among the thoracic manifestations of the CTDs, the interstitial diseases are perhaps the most intriguing and widely studied. In this regard, an important consideration is that almost all the patterns of idiopathic interstitial pneumonias (IIPs) (but, importantly, not their prevalences), are mirrored in CTD-related interstitial lung disease (ILD). The radiologic and histopathologic features of the IIPs have been well documented ( Table 2 ). Indeed, it may be argued that the true utility of HRCT is not in predicting the underlying CTD (because this invariably requires the integration of clinical, serologic, and radiologic data), but in identifying the likely pattern of interstitial pneumonia, which not only has prognostic implications but also may influence the decision to biopsy. In this regard, it is important to stress that overlapping HRCT appearances (particularly for usual interstitial pneumonia [UIP] and nonspecific interstitial pneumonia [NSIP] ) are not uncommon; predominant ground-glass opacification is the important feature of NSIP, whereas the cardinal findings in UIP are subpleural, basal reticulation with honeycombing. In the absence of honeycombing, attempting to make a radiologic diagnosis of “possible” UIP rather than NSIP relies on the observation of less extensive ground-glass attenuation compared with coarse reticulation with a subpleural, basal predominance, and the absence of the other features inconsistent with UIP. However, the use of such an interpretation algorithm can still result in misdiagnosis of biopsy-proven NSIP as UIP on HRCT, and vice-versa. It is also worth stating that inter-reader agreement for honeycombing is, at best, fair to moderate and that differentiating honeycombing from peripheral traction bronchiolectasis, on transaxial interspaced HRCT images, is not straightforward.



Table 1

Thoracic manifestations of the connective tissue disorders

























Compartment Manifestation
Airways Bronchial wall thickening, bronchiectasis, obliterative bronchiolitis
Lung parenchyma Interstitial lung disease: interstitial pneumonias (see Table 2 )
Airspace disease: diffuse alveolar damage, pulmonary hemorrhage
Necrobiotic nodules, infection, malignancy
Pulmonary vasculature Acute/chronic pulmonary thromboembolism, pulmonary hypertension
Pleura/pericardium Pleural/pericardial thickening, nodularity or effusion, pneumothorax
Mediastinum Esophageal dilatation/dysfunction, enlarged mediastinal lymph nodes
Thoracic musculature Muscle dysfunction leading to ventilatory impairment


Table 2

High-resolution computed tomography (HRCT) and histopathological characteristics of the interstitial pneumonias encountered in connective tissue diseases





























Pattern HRCT Features Histopathologic Features Remarks
Usual interstitial pneumonia (UIP)


  • Subpleural basal predominant reticulation



  • Honeycombing ± traction bronchiectasis



  • GGO usually inconspicuous




  • Marked fibrosis, architectural distortion



  • ± Subpleural/paraseptal predominant honeycombing



  • Patchy involvement



  • Fibroblastic foci



  • Only mild interstitial inflammatory infiltrate

Criteria for definite, possible, and inconsistent with UIP (on both HRCT and pathology) and probable UIP (on pathology only) have been defined
Nonspecific interstitial pneumonia (NSIP)


  • Bilateral GGO



  • Some basal and subpleural predominance



  • Fine reticulation/irregular linear opacities



  • ± Traction bronchiectasis/bronchiolectasis



  • Honeycombing sparse/absent




  • Varying amounts of interstitial inflammation and fibrosis



  • Uniform appearance

Honeycombing may become more prominent with progression
Katzenstein grades of pure cellular (grade I), mixed cellular and fibrotic (grade II), and fibrotic (grade III) recognized histopathologically
Organizing pneumonia (OP)


  • Patchy, multifocal consolidation Subpleural or peribronchial distribution



  • May be associated with GGO, perilobular pattern



  • ± Centrilobular nodules or masses




  • Intraluminal plugs of inflammatory debris: buds of granulation tissue and Masson bodies (whorls of fibroblasts and myofibroblasts in a connective matrix)



  • Predominantly within the alveolar ducts and surrounding alveoli



  • Mild interstitial inflammation

May coexist with an NSIP pattern
Lymphocytic interstitial pneumonia (LIP)


  • Ill-defined GGO and centrilobular nodules



  • Peribronchial/interlobular septal thickening



  • Thin-walled perivascular cysts



  • ± Lymph node enlargement




  • Diffuse interstitial infiltration: mostly T-lymphocytes, plasma cells, and macrophages



  • Predominantly alveolar septal distribution



  • Frequent bronchial mucosa-associated lymphoid tissue hyperplasia

Can be regarded as interstitial-predominant variant of diffuse pulmonary lymphoid hyperplasia
Follicular bronchiolitis is another form of predominantly peribronchial/peribronchiolar lymphocytic infiltrate seen in rheumatoid arthritis

Abbreviation: GGO, ground-glass opacity.

Adapted from Refs.


In addition to what has already been stated previously, some general comments about the imaging of CTD-related ILD also can be made. First, with the notable exception of RA (in which a UIP pattern is most prevalent ), NSIP is the dominant pattern of CTD-related ILD. Second, the presence of either mixed HRCT patterns (eg, features of lung fibrosis together with consolidation) or the coexistence of an ILD with, say, signs of pleural or airways disease, should alert the radiologist to the possibility of an underlying CTD, particularly when extrapulmonary abnormalities are present. Third, the notion that ground-glass opacification on HRCT denotes potentially reversible or treatable disease is no longer considered true for the idiopathic interstitial pneumonias or the CTD-related ILDs. Ground-glass opacification may indicate fine fibrosis, particularly when there is traction bronchiectasis/bronchiolectasis. The obvious message is that the mere presence of ground-glass attenuation should not be used as justification for potentially toxic therapy.




Imaging of individual connective tissue diseases


Rheumatoid Arthritis


The pulmonary and pleural manifestations of RA have continued to receive attention ever since the early report from 1948 by Ellman and Ball. The true prevalence of RA-associated thoracic disease is difficult to predict, chiefly because of differences in the populations studied and discordance between HRCT findings and symptoms. With respect to RA-related ILD, Gabbay and colleagues reported that, in a population of 36 patients with joint manifestations of recent (2 years) onset, one-third had HRCT manifestations of ILD, but there was clinically significant disease in only 14%. However, when clinically significant ILD was defined as RA-associated ILD that contributes to death, Olson and colleagues found that the prevalence of such ILD was closer to 7% in women and just less than 10% in men. The lifetime risk of developing ILD in patients with RA has been estimated at 7.7%. In another study of patients with RA having one or more of symptomatic ILD, abnormal pulmonary function tests, or chest radiographs, Biederer and colleagues found some form of HRCT abnormality in 92% of patients. As such, the prevalence of HRCT abnormalities is undeniably greater than the proportion with clinically relevant pulmonary disease.


Airway disease


In patients with RA, airway abnormalities are common. Bronchiectasis, with or without bronchial wall thickening, may be seen in up to 30% of patients ( Fig. 1 ). Peripheral branching nodular opacities (the tree-in-bud pattern) and heterogeneity in lung density (the so-called mosaic attenuation pattern), both indicating involvement of small airways, also are relatively frequently seen on HRCT. In one study of 50 patients with no evidence of RA-associated ILD, Perez and colleagues, identified bronchiectasis in 30%, with a third of such cases associated with bronchial wall thickening. It is also noteworthy that, with time, the extent of bronchiectasis and severity of bronchial thickening generally worsens with long-standing RA. The tree-in-bud pattern on HRCT indicates intraluminal/peribronchial small airway exudate and is reported in approximately 10% to 20% of patients with RA ( Fig. 2 ). Tree-in-bud opacities with centrilobular nodules also are seen in the rare entity of follicular bronchiolitis, which is characterized histologically by peribronchial lymphoid hyperplasia. Indirect evidence of small airways disease, reflected on HRCT by mosaic attenuation, is present in 44% to 52% of cases ( Fig. 3 ). Although patients with this HRCT pattern may be asymptomatic, a subset with the functional entity of obliterative bronchiolitis in RA, as first reported by Geddes and colleagues, experience rapidly progressive dyspnea and airflow obstruction.




Fig. 1


A 69-year-old man with a long-standing history of RA. HRCT demonstrates minimal bronchial wall thickening and cylindrical bronchiectasis, together with some nonspecific pleural thickening ( arrow ).



Fig. 2


A 41-year-old woman with a short history of RA and recent increasing cough. Transverse HRCT slice demonstrates quite marked centrilobular nodularity in the middle lobe and lingula, suspicious for an exudative small airways infection or follicular bronchiolitis. Mild cylindrical bronchiectasis and bronchial wall thickening are also noted.



Fig. 3


An 88-year-old man with long-standing RA. Cylindrical bronchiectasis is accompanied by mild mosaic attenuation on inspiration ( A ), accentuated on expiratory imaging ( B ), due to air-trapping.


Interstitial lung disease


It is now accepted that a pattern of UIP is more common than NSIP in patients with RA, despite earlier reports of equal or higher prevalence of NSIP. Although Tanaka and colleagues reported that HRCT findings accurately reflect histopathologic features, in most cases the concordance between histopathologic and HRCT diagnosis can vary. Indeed, Lee and colleagues demonstrated a histopathologic UIP pattern in 10 (56%) of 18 patients with RA, 9 of whom demonstrated HRCT-concordant UIP with subpleural reticulation and honeycombing ( Fig. 4 A). By way of contrast, a more recent study, which used the stricter HRCT definition of “definite UIP,” resulted in only 19 (45%) of 42 histopathologically proven cases of UIP being radiologically classified as UIP. Broadening the UIP definition to include cases of possible UIP (see Fig. 4 B) improved the HRCT sensitivity for a UIP diagnosis to 81%, but at the expense of a diminished specificity and poorer agreement between radiologists. The variation in radiologic-pathologic concordance is undoubtedly due to differences in the proportions of patients with honeycombing, but this variation cannot easily be explained by the duration of disease, as honeycombing may be present in equal proportions in both early and long-standing RA.




Fig. 4


HRCT for ILD in 2 patients with RA. ( A ) In a 73-year-old man with increasing breathlessness, a transverse HRCT slice demonstrates the definite UIP pattern of subpleural basal predominant honeycombing and traction bronchiectasis ( arrow ). ( B ) In a 74-year-old woman with recurrent cough, the HRCT demonstrates only minor subpleural reticulation and no definite traction bronchiectasis, but no features inconsistent with UIP. Honeycombing at the extreme base ( block arrow ) was thought to be present by one radiologist but not another at multidisciplinary discussion; a radiologic pattern of possible but not definite UIP was assigned.


There is an important possible link between smoking and pulmonary fibrosis in patients with RA. It would appear that tobacco smoke predisposes to RA-associated ILD, particularly UIP over NSIP, although the prognostic implications of smoking in RA-associated ILD are yet to be determined. Intriguingly, a recent study demonstrated that emphysema, in both smokers with RA-associated ILD and smokers with idiopathic pulmonary fibrosis (IPF), was significantly more prevalent than in controls who smoked but who did not have chronic obstructive pulmonary disease (48% and 35% respectively, vs 15%), despite a lower pack-year smoking history. This may be indirect evidence that shared mechanisms exist between smoking-related lung damage and RA-associated fibrosis.


In general, lung abnormalities in RA are either exclusively or predominantly present in the lower zones. It is also worth noting that, that although ILD and airway abnormalities may coexist, one or the other usually predominates. In this regard, one hypothesis is that variability in human leukcocyte antigen subtypes may predispose to bronchiectasis, rather than ILD, in RA.


Non–interstitial lung disease/airway manifestations of rheumatoid arthritis


Among the other recognized complications of RA in the chest, necrobiotic nodules (which have historically been considered a classic feature of RA-associated pulmonary involvement ) are surprisingly rare; the frequency and HRCT characteristics of necrobiotic nodules are difficult to ascertain because of a paucity of pathologic correlation. Pleural disease (usually pleural thickening rather than effusions), is common in RA, being present in up to 30% of patients on HRCT and 50% in one autopsy series. Enlarged mediastinal lymph nodes in RA also have been reported.


Systemic Sclerosis


Interstitial lung disease


ILD is prevalent in patients with SSc. Indeed, a higher prevalence of ILD on HRCT (reportedly between 36% and 91% ), is seen in SSc compared with the other CTDs. Pulmonary involvement contributes to mortality and ranks third in frequency behind cutaneous and peripheral vessel involvement in patients with SSc. Although the extent of ILD on HRCT may vary between cases, limited disease is more common: in one large study of more than 200 patients, the median extent of abnormal lung was only 13% (range = 1%–84%), reflecting, in part perhaps, the tendency for pulmonologists to “screen” patients for ILD.


It is now accepted that NSIP is more common than UIP as a pattern of ILD in SSc. Other patterns of interstitial involvement, such as organizing pneumonia (OP), are rare. In patients with SSc-associated NSIP, basal predominant subpleural ground-glass opacification with superimposed fine reticulation is the typical HRCT finding ( Figs. 5 and 6 ). Honeycombing may be present in up to 40% of cases and, interestingly, appears to be more common in limited SSc. Honeycombing tends to be limited in extent and characterized by a microcystic pattern. In one cohort of 52 patients with SSc with a median duration of 6.8 years of extrapulmonary disease, Remy-Jardin and colleagues showed that 32 patients had abnormalities on HRCT: ground-glass opacification and a nonseptal linear pattern were present in 26 (50%) and 6 (12%) cases, respectively. When honeycombing was present, it was localized in most (12 of 19 cases, 63%). The extent of honeycombing can increase on serial HRCT and has been shown to correlate with a decline in the diffusion capacity for carbon monoxide (DLco). From this, it is tempting to postulate that honeycombing is more prevalent in patients with disease of longer duration. However, no such relationship was observed in the study by Remy-Jardin and colleagues. In addition, the observation on follow-up CT that ground-glass attenuation not only fails to resolve with treatment, but is replaced by honeycombing, supports the notion that ground-glass opacities may represent fine fibrosis rather than a predominantly cellular inflammatory infiltrate.




Fig. 5


A 55-year-old woman with a new diagnosis of SSc, exertional breathlessness, and DLco. Transverse HRCT slice demonstrates patchy, quite extensive pure ground-glass opacity with no admixed reticulation or traction bronchiectasis.



Fig. 6


A 62-year-old woman with known SSc and increasing breathlessness over 18 months. Transverse HRCT just below the aortic arch ( A ) and at the bases ( B ) demonstrate ground-glass opacity with some admixed reticulation in the upper lobes, becoming more predominant in the lower lobes where there has been quite marked volume loss (note the retracted position of the left oblique fissure) and traction bronchiectasis ( arrow in B ). Traction bronchiectasis is a poor prognostic sign.


A preponderance of ground-glass opacity, fibrosis of limited coarseness, and the relative absence of honeycombing should allow the differentiation of SSc-associated NSIP from patients with IPF. However, it must be remembered that the coarseness of fibrosis and extent of ground-glass opacity (relative to reticulation) may be similar in patients with idiopathic and SSc-associated NSIP, making discrimination between the latter 2 entities impossible on parenchymal appearances alone. In practice, radiologists suspecting an HRCT pattern of NSIP may look for the other ancillary signs of diffuse SSc, such as esophageal dilatation or soft tissue calcinosis.


Non–interstitial lung disease manifestations of systemic sclerosis


Mediastinal lymph node enlargement, esophageal dilatation, and pleural abnormalities are the other CT abnormalities in patients with SSc. Nodal enlargement per se is seen in many diffuse lung diseases and has been correlated with the CT extent of ILD in SSc. Esophageal dilatation is present in 62% to 82% of patients, but whether or not esophageal dysmotility predisposes to, or correlates with, ILD is unclear. Pleural abnormalities, manifest as subpleural micronodules or pleural thickening, are reported in as many as 81% of cases at autopsy, but the frequency in HRCT studies is more variable. Finally, for the sake of completeness, mention also is made of the complication of pulmonary hypertension in SSc, which is a key contributor to mortality, and is discussed in greater detail later in this article.


Systemic lupus erythematosus


Lung disease in SLE is perhaps more prevalent than might first be appreciated; in one study, more than half of patients developed some feature of pleuroparenchymal involvement at some point during the course of disease. Autopsy series suggest that pleural and pericardial abnormalities predominate ; however, the reported prevalence, based on radiologic investigations, has varied. For instance, Sant and colleagues and Fenlon and colleagues reported pericardial thickening and pleural involvement (ie, pleural tags, thickening, and effusions) in only 15% and 17% of HRCTs, respectively. By way of stark contrast, Ooi and colleagues reported pleural disease in as many as 80%.


Interstitial lung disease


Interstitial lung involvement (characterized by thickened interlobular septa, and parenchymal and subpleural bands) has been seen in 33% to 38% of patients with SLE, but, when present, is usually mild in extent and severity and asymptomatic. Against this, the preponderance of honeycombing and architectural distortion in just over half of patients with an HRCT-derived ILD diagnosis studied by Ooi and colleagues is a salutary reminder of the influence of selection bias and small samples in many CTD-ILD studies. Indeed, it is believed that symptomatic ILD occurs in the minority (ie, <10%) of patients. A possible caveat is that a higher prevalence of interstitial abnormalities has been noted in patients with the antiphospholipid syndrome, in which irregular subpleural linear opacities, ground-glass opacity, reticulation, and interlobular septal thickening are seen. The more serious and potentially life-threatening complications of lupus pneumonitis and pulmonary hemorrhage are also recognized in SLE and are discussed in the section on pulmonary complications of CTD-associated disease.


An intriguing entity seen in patients with SLE that bears mentioning is “shrinking lung syndrome,” characterized by unexplained and often progressive dyspnea, physiologic small-volume lungs, and diaphragmatic elevation but with clear lungs on plain radiograph ( Fig. 7 ). This clinico-radiologic constellation has been observed in approximately 10% of patients with SLE. On HRCT there is bibasal atelectasis coupled with an elevated diaphragm. The pathogenesis of shrinking lungs remains unclear, but the role of primary diaphragmatic and/or respiratory muscle weakness (caused by a postulated subclinical myopathy ) is not supported by electrophysiologic data.




Fig. 7


A 38-year-old woman with SLE and increasing breathlessness. Chest radiograph demonstrates a raised left hemidiaphragm ( A ), whereas transverse HRCT on lung windows at the lung bases demonstrates the associated atelectasis at both bases in addition to the raised left hemidiaphragm ( B ), with some pericardial thickening as a sequelae of previous pericarditis also noted on mediastinal windows ( C ). The appearances were suspicious for the “shrinking lung” syndrome.


Polymyositis/dermatomyositis


In patients with PM/DM, lung disease can be problematic, and this is particularly true in the context of serum antibodies to aminoacyl-tRNA-synthetases (the so-called “antisynthetase” antibodies). Of the antisynthetase syndromes, antibodies to histidyl-transfer RNA (more conveniently called anti-Jo1) are the most prevalent, and their importance is that ILD reportedly occurs in more than 50% of patients with anti-Jo1 positivity.


The typical patterns of lung disease in PM/DM are NSIP and OP. Understandably, in most patients with ILD associated with PM/DM, the imaging features will reflect these patterns. Ground-glass opacification (present in 63%–92%) and consolidation (seen in 26%–55%) are frequently present, often with a peribronchovascular and lower lobe predilection ( Fig. 8 ). The consolidation and ground-glass opacification is often admixed with traction bronchiectasis and reticulation, corresponding to a histologic pattern of OP on a background of fibrotic NSIP. Indeed, the presence of predominantly basal reticulation and ground-glass opacification, traction bronchiectasis, volume loss, and scattered bronchocentric foci of consolidation may be the radiologic clue to the diagnosis of an antisynthetase syndrome. Foci of the perilobular pattern, a recognized manifestation of OP, also may be seen.




Fig. 8


A 31-year-old woman with a new diagnosis of polymyositis and anti-Jo1 antibody positivity. Transverse HRCT slices show patches of peripheral consolidation in the upper lobes ( A ) and more bronchocentric consolidation in the lower lobes ( B ), typical of the OP pattern seen in these patients.


There has been recent interest in the influence of different antibody profiles in DM and HRCT patterns. For instance, in clinically amyopathic dermatomyositis (CADM), the anti-CADM-140 antibody (recently renamed the anti-melanoma differentiation-associated gene 5 [anti-MDA5] antibody) is believed to be a factor in the development of ILD. Accordingly, in one small study, 12 of 25 patients with DM who were anti–CADM-140 positive, had significantly more lower-zone consolidation/ground-glass opacity and an absence of intralobular opacities than those who were negative for anti–CADM-140 ( Fig. 9 ). Hoshino and colleagues also found a strikingly higher frequency of ILD in anti-MDA5–positive subjects as compared with those without anti-MDA5 antibodies (95% vs 32%, respectively) and there was a trend toward rapidly progressive lung disease. An interesting, but as yet unexplained, observation is the increased frequency of pneumomediastinum in patients with DM with a CADM phenotype, especially in patients who are anti-MDA5 positive (see Fig. 9 ).




Fig. 9


A 50-year-old woman with a recent diagnosis of clinically amyopathic dermatomyositis and anti-MDA5 antibody positivity. ( A ) Transverse conventional (non–high-resolution) CT (2-mm slice thickness) demonstrates a bandlike opacity of consolidation in the apical segment of the right lower lobe, typical of OP. Note the pneumomediastinum ( arrow ), well-described in such patients. ( B ) Follow-up HRCT slice 3 months later shows persistence of consolidation in the right lower lobe, with the perilobular pattern of OP ( block arrow ) noted. The pneumomediastinum has resolved.


Sjögren syndrome


The prevalence of HRCT abnormalities in patients with SS with pulmonary symptoms has been estimated at 11%, even though, somewhat counterintuitively, more than two-thirds of asymptomatic patients may demonstrate such abnormalities. In SS, patients are more liable to develop impairment of upper respiratory tract immune defenses, glandular dysfunction, and bronchiolar inflammation, which probably explains why signs of airway involvement (ie, bronchial wall thickening, dilatation, and a mosaic attenuation pattern) are commonly seen. The presence of airways disease makes it difficult to gauge exactly how much of the parenchymal injury seen in SS is a consequence of airways-based inflammation, rather than a direct “hit” on the lung parenchyma. This notwithstanding, the ILD in SS is most commonly of an NSIP pattern, characterized by lower-zone ground-glass opacification and interlobular septal thickening.


A pattern of lymphoid interstitial pneumonia (LIP) has been considered the commoner SS-associated ILD, but it is possible that such statements were based on the subjective histologic interpretation of a homogeneous lymphocytic infiltrate as LIP rather than a cellular NSIP. In some patients with LIP related to SS, HRCT may show patchy ground-glass opacities together with thin-walled parenchymal cysts, sometimes measuring up to 6 cm in diameter, as well as peribronchovascular, centrilobular, and subpleural nodules ( Fig. 10 ). The association among SS, LIP, and other lymphoproliferative disease is also worth remembering. Indeed, the spectrum of lymphoproliferative disorder in SS varies and ranges from benign, as exemplified by the entity of follicular bronchiolitis (manifesting as a predominant tree-in-bud pattern on HRCT), to overt malignant lymphoma (usually of non-Hodgkin type). Lymphoma in patients with SS will present as nodules larger than 10 mm in diameter, foci of consolidation, mediastinal lymph node enlargement, and pleural effusions. A rare but striking presentation of LIP in SS occurs when there is associated pulmonary amyloid deposition; on HRCT there are multiple bizarrely shaped nodules (which may appear calcified) often adjacent to cysts, with no particular zonal predilection.




Fig. 10


A 67-year-old woman with known Sjögren syndrome and biopsy-proven LIP. Transverse HRCT slices show multiple foci of bronchocentric ground-glass opacity ( A ), as well as multiple thin-walled parenchymal cysts ( block arrow in B ). Traction bronchiectasis is present ( arrowhead in B ), concerning for an adverse prognosis. Note the surgical suture material postbiopsy in the posterior right upper lobe ( arrow in A ).


Mixed connective tissue disease


There is continuing debate about whether MCTD represents a distinct entity or a simply an “overlap syndrome” with features of, most commonly, SSc, SLE, and PM/DM. There is support for the latter hypothesis in that the thoracic manifestations of MCTD generally represent an overlap of features found in the other CTDs listed previously. Lung disease, based on review of HRCT, is reported in more than 50% of patients with MCTD. The most common abnormalities are reticulation and ground-glass opacification, which tend to be basal predominant and typically of limited extent ( Fig. 11 ). In one study, there was a reticular pattern at HRCT in just more than one-third of patients; in most cases, reticulation was admixed with fine intralobular lines or cysts measuring smaller than 4 mm in diameter.




Fig. 11


A 52-year-old woman with MCTD with overlapping clinical features of SSc and RA. Transverse CT slice at the lung apices ( A ) demonstrates mild subpleural reticulation, in addition to the prominent dilated esophagus and marked soft tissue calcinosis surrounding the sternoclavicular and glenohumeral joints, while marked traction bronchiectasis and ground-glass opacity is seen at the bases ( B ), compatible with a fibrotic NSIP.


The reported prevalence of certain abnormalities in MCTD may differ from its “component” CTDs. In a rare direct comparison, Saito and colleagues observed a lower frequency of ground-glass opacity in patients with MCTD compared with patients with SSc, SLE, and PM/DM, but found that interlobular septal thickening was present in all patients with MCTD, with an unexpectedly high proportion of honeycombing. A striking preponderance (98%) of subpleural nodularity also was noted in another study, the reasons for which are unclear. Interestingly, although the clinical features of PM/DM may be present in patients with MCTD, the mixed NSIP-OP pattern described in PM/DM has, to the best of our knowledge, not been reported in MCTD. Other radiologic findings of MCTD-related thoracic disease include pleural thickening/effusions, esophageal dysmotility, centrilobular nodularity, and airway involvement, but are rare.


Pulmonary Complications in Connective Tissue–Associated Disease


As discussed previously, the lung complications of CTDs are reasonably common. For practical purposes, these can be broadly thought of as acute and nonacute sequelae ( Table 3 ). The imaging implications of these complications are reviewed in the following sections.


Sep 28, 2017 | Posted by in RHEUMATOLOGY | Comments Off on Imaging of Pulmonary Involvement in Rheumatic Disease

Full access? Get Clinical Tree

Get Clinical Tree app for offline access