Thoracic manifestations of IgG4‐related disease

INTRODUCTION AND HISTORY

Immunoglobulin G4‐related disease (IgG4‐RD) is a recently described systemic fibroinflammatory disease characterized by unifying histopathological findings. ¹ The disease can affect virtually any organ, either singly or in association, but most commonly involves lymph nodes, liver, pancreas, lacrimal and salivary glands. ²

The description of the first cases of IgG4‐RD dates from 1961: Sarles et al. reported chronic autoimmune pancreatitis in patients with a distinctive histological feature of fibrosis and lymphoplasmacytic infiltration. ³ In 2001, Hamano et al. reported fast‐moving band in the beta‐gamma region of the serum protein electrophoresis of patients with sclerosing pancreatitis, representing elevated IgG4 levels. ⁴ In 2003, abundant polyclonal IgG4+ plasma cells were found within polyclonal lymphoplasmacytic infiltrates in tissue samples from patients with autoimmune pancreatitis, both in the pancreas and in the liver and gallbladder. ⁵ Based on these findings, in 2006, Kamisawa proposed the existence of an IgG4‐related sclerosing disease corresponding to pancreatic and extrapancreatic fibroinflammatory lesions rich in IgG4. ⁶ The disease has been variously named, including IgG4‐related autoimmune disease, IgG4‐associated multifocal systemic fibrosis, IgG4‐related systemic disease, IgG4‐related sclerosing disease or IgG4 plasmocytic syndrome. In 2011, an international group of investigators agreed on a uniform nomenclature and diagnostic criteria for this entity consensually referred to as IgG4‐related disease. ⁷ This term emphasized the systemic infiltration of IgG4+ plasma cells in the organs involved and the frequency of high serum IgG4 levels without implying any pathogenic role of IgG4. This new nosological entity included many syndromes previously considered independent and idiopathic (inflammatory pseudotumours, Mikulicz syndrome, Küttner tumour, Riedel thyroiditis, mediastinal and retroperitoneal fibrosis).

Today, the use of the term IgG4‐RD has reached a consensus and is based on precise classification criteria published in 2019. ⁸ It is a polymorphic systemic disease, whose most common presentations are hypertrophies of the salivary and lacrimal glands, lymphadenopathy, pancreatitis, sclerosing cholangitis, retroperitoneal fibrosis and tubulointerstitial nephritis. ⁹ This wide heterogeneity has recently led to the description of four subgroups of patients (hepatobiliary, disease limited to the head and neck, Mikulicz and retroperitoneal fibrosis), whose pathophysiological, prognostic and therapeutic response particularities remain to be evaluated, ¹⁰ but highlight the mainly abdominal and ENT tropism of the disease.

As a matter, the thoracic involvement of IgG4‐RD has been only recently described. ¹¹ The disease can affect nearly all thoracic organs, including heart, vessels, lungs, pleurae, lymph nodes, retromediastinum and bones. This thoracic heterogeneity has long been poorly characterized ¹² but is the subject of a growing number of publications. Here we review the current knowledge about pathogenesis, epidemiology, diagnostic strategies and therapeutic modalities of thoracic IgG4‐RD.

PATHOGENESIS

EPIDEMIOLOGY

To date, the incidence of the disease has been estimated in Japan to be approximately 1 per 100,000 ³⁹ and prevalence has been estimated to be 0.8/100,000. ⁴⁰ The disease occurs predominantly in adults, with a median age at diagnosis of 60 years, and affects mainly men (1 man for 0.77 women). The age of onset, usually after 50 years, and the male predilection are unusual epidemiological parameters for an autoimmune disease.

Like IgG4‐RD, thoracic involvement is typically seen in middle‐aged to elderly adults, more commonly in men than women. ⁴¹ It is estimated that 15 and 35% of patients with IgG4‐RD are affected. ¹⁰ , ⁴² , ⁴³ , ⁴⁴ Most cases of thoracic involvement with IgG4‐RD are found incidentally during the workup of extrathoracic lesions. ⁴² The most common thoracic disorders seem to be bronchopulmonary disease and lymph node involvement. ⁴⁴

Isolated thoracic involvement was estimated to occur in 13% of patients. ⁴² , ⁴⁴ Isolated lung involvement was estimated at 8% in a Japanese study based on 4304 patients. ³⁹

CLINICAL PRESENTATION

IgG4‐RD usually manifests subacutely. More than half of the patients are asymptomatic according to several studies. ¹¹ , ⁴² , ⁴⁴ Two small studies totaling 53 patients with systemic IgG4‐RD have detailed the general clinical signs present at diagnosis. ⁴⁵ , ⁴⁶ The main general symptoms included asthenia in 14 patients (25%), weight loss in 11 (20%) and fever in 4 (8%). Clinical signs of thoracic involvement could be present in more than 10% of the cases. They are nonspecific and depend on the location of the lesion, including cough, dyspnoea, haemoptysis, asthma or chest pain. ⁴⁷

IMAGING PRESENTATION

Chest CT‐scan is the most widely performed imaging method to evaluate IgG4‐RD thoracic involvement. Chest CT‐scan presentation of IgG4‐RD is polymorphic and non‐specific. ¹¹ , ⁴² , ⁴⁸ , ⁴⁹ Ideally, patients should be evaluated with dedicated thoracic imaging, including thin‐section CT scans and high‐resolution reconstruction filters. Chest CT scans should then be analysed by radiologists specialized in thoracic imaging and familiar with the disease. In our experience, specialized review in a tertiary centre significantly improved the number of abnormalities diagnosed. This was particularly pronounced in the case of pulmonary involvement, especially for discrete images of round shaped ground‐glass opacities (GGO) or fine reticulations. ⁴⁴

Through literature, seven main patterns of thoracic IgG4‐RD involvement are described ¹¹ and can be associated. Four patterns affect lungs: nodular, GGO, interstitial disease and peribronchovascular (Figure 1). The three others correspond to chest extrapulmonary lesions: lymph node, pleural and retromediastinal patterns (Figure 2). In the framework of the latter, the presence of a paravertebral soft band is relatively specific of the disease. This corresponds to a thickening of the paravertebral sulcus, in the lower thoracic region involving two or more vertebrae in a row. This sign is mostly located on the right side (97%, n = 31). ⁵⁰ Other CT‐scan findings than those included in these seven patterns are possible, but much rarer.

The relative frequency of these different patterns fluctuates between studies. In a cohort of 48 patients with thoracic involvement, we found that peribronchovascular involvement was the most frequent, affecting 60% of those patients. Fujinaga et al. ⁵¹ identified hilar lymph node involvement as the most prevalent (80%, n = 90). The GGO pattern was the most common according to Inoue et al. ¹¹ (70%, n = 13), and Seraya et al. ⁴³ (70%, n = 16), whereas it was the nodular pattern in the study by Kang et al. (30%, n = 37). ⁵² None of these patterns is specific for IgG4‐RD. These discrepancies could be partly explained by different patient recruitment methods, bias related to retrospective data collection, the lack of central imagery reviewing and different pattern definitions from one study to another.

Even in the setting of well‐known IgG4‐RD, any chest imaging abnormalities are not necessarily secondary to IgG4‐RD. Thus, classification criteria for thoracic involvement have been developed by Corcoran et al. to assess the responsibility of the disease in observed thoracic lesions. ⁴² According to their criteria, thoracic involvement is categorized between ‘definite’ (biopsy‐proved), ‘highly probable’ (typical imaging finding with no alternative explanation and response to treatment), ‘probable’ (typical radiology with no alternative explanation) and ‘possible’ (radiology consistent but plausible alternative cause identified). In the absence of biopsy or evaluation of the effect of a test treatment, the association of two or more thoracic patterns in the same patient could suggest IgG4‐RD.

Some thoracic imaging patterns might be associated with specific IgG4‐RD extra‐thoracic involvements. For example, our previous study of 48 patients with IgG4‐RD thoracic involvement suggested that round‐shaped GGO may be associated with a more frequent IgG4‐related pancreatitis, that lymph node enlargement might be associated with a greater frequency of sialadenitis, and that interstitial disease could be associated with a more frequent hypereosinophilia. ⁴⁴ Underlying pathophysiological mechanisms are not established, and these data need to be confirmed by further studies.

¹⁸F‐FDG PET/CT can also be used for a comprehensive assessment of organ involvement. It could have greater sensitivity for detecting disease in arteries (aortitis and periaortitis), salivary glands and lymph nodes. ^{53 18}F‐FDG PET/CT can be useful detecting clinically asymptomatic IgG4‐RD organ involvement in the thorax, then selecting the most diagnostically informative biopsy site, assessing response to therapy and monitoring for disease relapse. ⁵⁴

BIOLOGICAL PRESENTATION

Biologically, two key markers for diagnosis and follow‐up of the disease have been identified to date: serum IgG4 level and detection of circulating plasmablasts.

Elevated serum IgG4 concentration (>135 mg/dl) represents an important diagnostic consideration. This threshold is part of all classification and diagnosis criteria, including the last ACR‐EULAR criteria published in 2019, but exhibits inconstant sensibility and specificity and so a low positive predictive value (34%) for the diagnosis of IgG4‐RD. ⁵⁵ Overall, it is estimated that 60% of patients with IgG4‐RD have a serum IgG4 level greater than 135 mg/dl, but this percentage varies according to ethnicity and the number of organs affected. Caucasian patients cohorts report less than 50% elevation in IgG4, ⁵⁶ compared to over 90% in Asian cohorts, ⁵⁷ and patients with multiple organ damage have significantly higher IgG4 levels. ⁵⁸ Markedly elevated serum IgG4 is rather more specific for IgG4‐RD: a 5 g/L cut off is associated with a specificity of 90%. ⁵⁹ Of note, IgG subclasses are most often measured by immunonephelometry, a technique that may show limitations due to prozone effect and interactions of IgG4 with measurement of IgG1 and IgG2, which is why some authors have proposed the preferential use of mass spectrometry in IgG4‐RD. ⁶⁰

Flow cytometric detection of circulating plasmablasts is also a useful biomarker for diagnosing IgG4‐RD. A 900/ml cut‐off could be associated with reported sensitivity more than 90% and specificity of 80%. ¹³ This biomarker was found to correlate with disease activity and the number of affected organs, independently of serum IgG4 levels. ⁶¹ The levels of other circulating cells identified during IgG4‐RD (CD4+ CTL, Tfh, B memory cells) are correlated with disease activity but clear thresholds between healthy and diseased individuals remain to be determined.

Moreover, a nonspecific biological inflammatory syndrome is often noted. Elevated C‐reactive protein levels are found in a quarter of patients, but this increase is usually modest. ⁶² Polyclonal hypergammaglobulinemia (total serum IgG levels >1800 mg/dl) is reported in 60% of the patients, which may be associated with high IgG2 and low IgA levels. ⁶³ Other non‐specific biologic markers observed include hypocomplementemia (40%), high soluble IL‐2R levels, positivity of antinuclear antibody (30%) or rheumatoid factor (20%), eosinophilia (30%) and elevated IgE (60%), especially in patients with asthma and atopy. The combination of asthma, hypereosinophilia and elevated IgE is not uncommon in IgG4‐RD and seems to be correlated with the risk of disease recurrence. ⁶⁴ However, at the time of diagnosis, this association should raise the possibility of atopic asthma, hypereosinophilic syndrome (HES), ANCA vasculitis or immunoallergic aspergillosis. For this reason, hypereosinophilia greater than 3G/L constitutes an exclusion factor for the disease according to the ACR/EULAR classification criteria.

Many other markers, that we not detailed in this review, are currently under investigation in IgG4‐RD. ⁶³ Summary of blood biomarkers used for diagnosis, assessment of disease activity and prediction of relapse risk is presented in Table 1.

HISTOPATHOLOGY OF THORACIC LESIONS

IgG4‐RD demonstrates common histologic features in most of organs that may be affected. The three main histopathological characteristics of IgG4‐RD are the presence of a polyclonal lymphoplasmacytic infiltrate, a storiform fibrosis and obliterating phlebitis. ²

The dense lymphoplasmacytic infiltrate is composed of T cells, B cells, occasional germinal centres and plasma cells. It can cause obliteration of the venous channels, leading to obliterative phlebitis. An essential feature of this infiltrate is the abundant presence of IgG4 plasma cells, which should be assessed semi‐quantitatively by immunostaining of the total number of IgG4 cells per high‐power field (/hpf, whose positivity thresholds depend on the site of involvement) and measurement of the IgG4/IgG plasma cell ratio (generally considered abnormal if the ratio is >40%). ⁷ Storiform fibrosis is a swirling, ‘cartwheel’ pattern of fibrosis sometimes with a patchy distribution that may be missed with small biopsies. Other ‘minor’ histopathological features include increased number of eosinophil and phlebitis without obliteration of the lumen. Absence of granulomatous inflammation and prominent neutrophilic infiltrate is required. Histological findings are neither specific nor sensitive and may lead to under or overdiagnosis. ⁶⁹ This resulted in a probabilistic histologic classification of the disease into three groups: ‘histologically highly suggestive of IgG4‐RD’ (presence of at least two of the three characteristic histological features), ‘probable histologic features of IgG4‐RD’ (only one of them) and ‘insufficient histopathologic evidence of IgG4‐RD’ (none of them), as it may be the case after previous therapy or progression to a fibrotic stage. ⁷

At the thoracic level, different organs can be investigated, with few histological specificities. Regarding lymphadenopathies, five morphological subtypes have been described, ⁶⁹ but the specificity of these histologic changes in the absence of other evidence of IgG4‐RD remains controversial: (a) multi‐centric Castleman disease‐like, (b) reactive follicular hyperplasia, (c) interfollicular expansion, (d) progressive transformation of germinal centre‐like and (e) inflammatory pseudotumour‐like. The most specific subtype is the last mentioned, but reactive follicular hyperplasia is most frequently observed. ⁷⁰ Fibrosis is only seen in inflammatory pseudotumour‐like lesions, and obliterative phlebitis is rarely observed. ⁷ IgG4+/plasma cells threshold is >50 in lymphadenopathies. Considering the variability of histologic patterns, the relative rarity of storiform fibrosis and obliterating phlebitis and the poor specificity of increased IgG4+ plasma cells in lymphadenopathies, the main value of lymph node biopsy is often to rule out differential diagnoses, such as lymphoma.

Regarding the lungs, histopathology of lesions shows diffuse lymphoplasmacytic infiltration, obliterative vascular changes and fibrosis with occasional eosinophilic infiltration. ⁷¹ The most frequent histopathological pattern is a lymphangitic distribution of inflammatory infiltrates rich in plasma cells. Several specificities should be noted. Arteritis characterized by a lymphoplasmacytic infiltrate is regularly observed and may lead to obliteration. ⁷² This obliterating arteritis is more frequent than obliterating phlebitis, unlike in other organs. ⁷ In addition, the characteristic storiform fibrosis seen in other organs is rarely found into pulmonary lesions. ⁷³ Consensus thresholds are >50 IgG4/plasma cells (/hpf) for surgical biopsies but >20 IgG4/ plasma cells (/hpf) for nonsurgical biopsies. ⁷ IgG4/IgG ratio threshold remains the same (>40%) for both types of biopsies.

Regarding the pleura, histological patterns correspond to those commonly observed in most of the other affected organs, with a threshold of >50 IgG4/plasma cells (/hpf). ⁷⁴ In the mediastinum, histopathology of lesions exhibits storiform cell‐rich fibrosis, lymphoplasmacytic infiltrate with IgG4‐positive plasma cells and obliterative phlebitis, ⁷⁵ which are not specific, and may be observed, for example, in histoplasmosis. ⁷⁶ Consensus statement on the pathology of IgG4‐related disease published in 2012 did not propose a specific IgG4/plasma cell threshold for fibrosis mediastinitis. The one suggested proposed for retroperitoneum (30IgG4/plasma cells) is therefore most often retained by analogy.

Several procedures are available to collect a sample of involved tissue, depending on the organ concerned, the location of the lesion, the patient's medical history and the experience of the centre. Mediastinal lymph nodes can be reached by endobronchial ultrasound (EBUS‐TBNA), surgical sampling or mediastinoscopy, while pleural involvement can be reached by percutaneous core cutting needle or surgical sampling. ⁴⁴ There are no studies comparing the cost‐effectiveness of these different techniques in the specific setting of IgG4‐RD. Nevertheless, given the relevance of lymphadenopathies architecture analysis in the diagnosis of IgG4‐RD, biopsy or surgical excision should be preferred to EBUS‐TBNA. If thoracic affected organs are not accessible for biopsy, a minor salivary gland biopsy may be performed, even without clinical signs of salivary involvement. Histologic specificities include possible conspicuous lymphoid follicle formation, and sometimes lacks obliterative phlebitis and storiform fibrosis. Moriyama et al. reported a sensitivity of 55% and a specificity of 100% for labial salivary gland biopsy involving 66 patients with suspected IgG4‐RD. ⁷⁷

DIAGNOSTIC OF THORACIC IGG4‐RD

The diagnosis of thoracic involvement of IgG4RD is a challenging process based on the agreement of clinical, biological, imaging and histological findings. Virtually, all intrathoracic organs can be affected (Table 2) but no feature is pathognomonic for the diagnosis. The diagnostic approach must therefore always consider screening for differential diagnoses before retaining the diagnosis, especially when the disease affects a single organ and is either identified incidentally on radiological studies or diagnosed unexpectedly in pathological specimens.

Clinically, the presence of extra‐thoracic organ involvement, especially in organs frequently affected by IgG4‐RD (pancreatitis, sclerosing cholangitis, lymphadenopathy, interstitial nephritis, sialadenitis and retroperitoneal fibrosis), supports the diagnosis. However, thoracic involvement may be isolated. ⁴² , ⁴⁴ Increased serum IgG4 levels support the diagnosis, but may be normal in IgG4‐RD while elevated in other conditions such as recurrent infections, cancers or autoimmune diseases. ⁷⁹ Histological analysis of at least one affected organ is most often required for the diagnosis. But here again, histology is susceptible to abuse: some conditions such as Castleman's disease or Rosai–Dorfman's disease may be responsible for IgG4‐RD‐like features. ¹² , ⁴⁷ Furthermore, HES and eosinophilic granulomatosis and polyangiitis (mostly ANCA‐negative phenotype) may be responsible for genuine overlapping syndromes with IgG4‐RD, defining a spectrum of eosinophil‐mediated diseases that is still poorly characterized. ²⁸ , ⁸⁰ , ⁸¹

The diagnostic approach (Figure 3) therefore consists of collecting all the clinical, biological, imaging and histological elements in favour of the diagnosis, while keeping in mind that each of these elements must raise specific differential diagnoses (Table 3). Pragmatically, patients who present with classic clinical, laboratory and radiological manifestations of IgG4‐RD but in whom biopsies are not feasible or noncontributory are generally given a working diagnosis of ‘suspected IgG4‐RD’ and treated as such, provided that sufficient efforts have been made to exclude IgG4‐RD mimics. ⁸² Among IgG4‐RD mimetics, sarcoidosis (interstitial lung disease, adenopathy, granulomatous sialadenitis, increased serum CEA levels), ANCA vasculitis (interstitial lung disease, sinusitis, hypereosinophilia, glomerulonephritis, anti MPO or PR3 antibodies), neoplasia (altered general condition, adenopathy, pulmonary nodules or carcinomatous lymphangitis) and tuberculosis (pulmonary nodules, adenopathy and extra‐pulmonary involvement) must be considered systematically in case of pulmonary involvement.

CLASSIFICATION AND STAGING

Classification and diagnosis criteria for IgG4‐RD have been published periodically since the recognition of the disease.

The most recent classifications are the ACR/EULAR criteria, published in 2019. ⁸ They are based on a score in which clinical, biological, imaging and histological findings are considered. Thoracic involvement is included through ‘pulmonary involvement’, in which only the aspects of peribronchovascular and septal thickening are recognized (+4 points), and ‘retromediastinal involvement’ (paravertebral band‐like soft tissue in the thorax, +10 points). A score of 20 or more classifies patients as having IgG4‐RD.

In 2017, comprehensive diagnostic criteria were published by Umehara et al., revised in 2021. ⁸³ They are based on the combination of CT‐objectified thoracic organ involvement, IgG4 levels above 135 mg/dl, and IgG4+ >10/HPF cell infiltration associated with a ratio of IgG4+/IgG >40%. According to this work, the presence of these three parameters characterizes the ‘defined’ disorders.

In order to standardize the criteria for evaluating response to treatment, which may vary from one specialty to another, an IgG4‐RD Responder Index was developed in 2012. ⁸⁴ This score has been shown to have a good correlation with the physician global assessment. ⁸⁵ Disease activity in 14 organs is rated from 0 to 4. The degree of urgency of damages and sequelae are also recorded, and serum IgG4 level are included in the scoring. Within thoracic involvement, lung, lymphadenopathy, large vessel including aorta and mediastinum are especially mentioned. This score is now used in randomized clinical trials to evaluate response to treatments and its follow‐up over time.

TREATMENT

Treatment of all symptomatic patients is recommended. Conversely, asymptomatic patients, such as those with isolated chest adenopathy, may remain untreated. Treatment of localized forms of the disease, such as a pulmonary nodule or inflammatory pseudotumour, may be limited to surgical removal. In other cases, the treatment of IgG4‐RD thoracic disease is based on immunosuppressive drugs.

International consensus statement on the treatment of IgG4‐RD has been published in 2015, ⁸⁶ based on expert opinion and review of available evidence. Steroids are suggested as a first line treatment, at 0.6 mg/kg. Efficacy is usually observed within the first 2 weeks of treatment. ⁸⁷ After 2–4 weeks of treatment at 0.6 mg/kg, steroids are usually tapered off over 3–6 months. Non‐response to steroids is uncommon (less than 5% of patients) and should lead to consideration of alternative diagnoses to IgG4‐RD. On the contrary, relapse following steroid reduction is quite common. It is estimated to occur in 60%–80% of patients. ⁸⁸ Steroid dependence may require the prescription of a second immunosuppressive drug to minimize the side effects of long‐term steroid use. It may also require the maintenance of very low doses of steroids over time (e.g., 5 mg/day), which decrease relapse rates compared to placebo. ⁸⁹

Rituximab is an anti‐CD20 monoclonal antibody that has been shown to be effective in reducing steroid dosage in IgG4‐RD patients with steroid dependence. The treatment is usually prescribed at the dose of 1 g on Day 1 and then on Day 15. Its effectiveness is generally maintained for several months. In the study by Carruther et al., ⁹⁰ 97% of patients (29 out 30) experienced a favourable response to Rituximab, and 40% maintained complete remission at 1 year. Ebbo et al. reported 93.5% of clinical response (29 out 31 patients), with 42% of relapse at 2 years. ⁹¹ Maintenance therapy at a dose of 1 g every 6 months is possible to limit the risk of late relapse or to control the onset of disease recurrence. Secondary hypogammaglobulinemia may then occur (3 patients out 31 Ebbo et al.), possibly associated with secondary infections, mainly affecting of the respiratory tract.

Disease‐modifying anti‐rheumatic drugs (azathioprine, mycophenolate mofetil, ciclosporin, methotrexate, …) are not very effective for induction of remission but may be used for steroids sparing. ⁸⁶ , ⁹² Because of the lack of prospective clinical trials, the international consensus guideline on the treatment of IgG4‐RD had only 50% agreement among experts on whether disease‐modifying antirheumatic drugs should be started early or not. ⁸⁶

Several emerging therapies offer promise. A phase 2 clinical trial of anti‐CD19 antibody inebilizumab showed an interesting outcome in 12 patients (decrease in IgG‐RD Responder Index score by ≥2 points). ⁹³ The use of lymphoma chemoimmunotherapy regimens such as fludarabine or bendamustine in combination with rituximab has been successfully reported for steroid and rituximab‐refractory cases. ⁵⁹ The use of dupilumab, an anti‐IL4/IL‐13 monoclonal antibody prescribed to treat asthma, seems interesting given the involvement of these two cytokines in the pathophysiology of the disease, but its efficacy has only been reported in isolated cases and remains controversial. ⁹⁴ , ⁹⁵ Abatacept, a CTLA4‐Fc fusion protein, has also been evaluated in a few case report and a proof of concept study. ⁹⁶ , ⁹⁷ Studies are ongoing to evaluate the effectiveness of rilzabrutinib (BTK inhibitor) and elotozumab (anti‐SLAM7). Anti‐CD38 (daratumumab) or proteasome inhibitors (bortezomib) also offer interesting perspectives to specifically target B cells. Anti‐IgE (omalizumab) or anti‐IL5 (mepolizumab) treatment may be options for patients with atopy or severe asthma with high IgE levels.

CONCLUSION

IgG4‐RD is a fibroinflammatory disease that can affect virtually all organs, mainly occurring in middle‐aged to elderly male patients. In the thorax, the disease can occur in all compartments (mediastinum, airways, lungs and pleurae), most often in association with extra‐thoracic manifestations. Clinical manifestations are nonspecific and usually lacking. CT‐scan findings can be divided into seven distinctive patterns, whose prognostic characteristics have yet to be determined. Diagnosis is based on a combination of clinical, biological and histological findings, as no marker is specific for the disease. Diagnostic approach must always consider screening for differential diagnoses especially when the disease affects a single organ, is identified incidentally on radiological studies, or is diagnosed unexpectedly in pathological specimens. Thoracic IgG4‐RD is generally highly responsive to immunosuppressive therapy with steroids and/or rituximab, but relapse is common. Further studies are needed to better characterize thoracic involvement of the disease, its triggers, pathophysiology and prognostic factors to elucidate more effective and specific therapies.

AUTHOR CONTRIBUTION

Romain Muller, Mikael Ebbo and Nicolas Schleinitz conceived and designed the work. Romain Muller and Mikael Ebbo drafted the article. Paul Habert, Laurent Daniel, Antoine Briantais, Pascal Chanez, Jean Yves Gaubert and Nicolas Schleinitz critically reviewed the manuscript.

CONFLICT OF INTEREST

None declared.

Muller R, Ebbo M, Habert P, Daniel L, Briantais A, Chanez P, et al. Thoracic manifestations of IgG4‐related disease. Respirology. 2023;28(2):120–131. 10.1111/resp.14422

DATA AVAILABILITY STATEMENT

The data that support this manuscript are available from the corresponding author, (Nicolas Schleinitz), upon request.