Skip to main content
The British Journal of Radiology logoLink to The British Journal of Radiology
. 2019 Nov 21;92(1104):20190473. doi: 10.1259/bjr.20190473

Cardiothoracic manifestations of Erdheim-Chester disease

Jeeban Paul Das 1,, Lola Xie 1, Chris C Riedl 1, Sara A Hayes 1, Michelle S Ginsberg 1, Darragh F Halpenny 1
PMCID: PMC6913371  PMID: 31386554

Abstract

Erdheim-Chester disease (ECD) is a rare non-Langerhans cell histiocytosis which can have a broad range of clinical and radiological presentations. Typically, ECD affects multiple organ systems, with skeletal involvement present in almost all ECD patients and cardiothoracic manifestations in more than half. Cardiac and thoracic involvement contributes significantly to morbidity and mortality in affected patients and may have prognostic implications. The diagnosis of ECD can be challenging due to its rarity and similarity to other systemic disease processes. Although the diagnosis can be suggested on imaging, histopathology and immunohistochemistry are required for confirmation. We describe the multimodal imaging features of mediastinal, cardiac, pleural and lung parenchymal ECD. This review identifies the most common radiological manifestations of cardiac and thoracic ECD on contrast-enhanced CT, fluorine18-fludeoxyglucose positron emission tomography/CT and cardiac MRI, and highlights the role of these cross-sectional techniques in disease diagnosis.

Introduction

Erdheim-Chester disease (ECD) is a rare multiorgan disease characterized histopathologically by infiltration of affected organs by foamy histiocytes,1,2 causing granulomatosis and fibrosis.3 Recognition and reporting of this uncommon pathology is increasing, however, it remains very rare with an unknown global incidence.4

Lesions in the skeletal system, kidneys, retroperitoneum and central nervous system are the most characteristic non-thoracic manifestations of ECD. Although a detailed discussion of all systemic radiological features of ECD is beyond the scope of this review, we note that the most frequently encountered manifestations of ECD in daily clinical practice include; symmetric disphyseal and metaphyseal osteosclerosis of the femora on plain radiography or CT, present in over 90% of patients, dense perinephric fat infiltration on cross-sectional imaging ("hairy kidney" sign), a feature in over two-thirds of patients and central nervous system ECD involvement seen in 25–50% of patients, including orbital mass lesions as well as thickening and enhancement of the pachymeninges and pituitary infundibulum.3

Increased frequency of thoracic cross-sectional imaging in ECD patients and advances in imaging quality have led to improved detection of cardiothoracic findings in this disease, with cardiothoracic involvement seen in 50–90% of patients.5–8 Identifying cardiothoracic ECD is important, as it denotes a worse prognosis with up to 40% of ECD-related deaths occurring due to cardiorespiratory involvement.1 Cardiothoracic ECD can affect the heart, mediastinum, pleura or lung parenchyma. Consequently, clinical presentations are variable, but include the sequalae of cardiac valvular disease, arrhythmia, myocardial infarction and interstitial lung disease.1,4–8

This pictorial review presents the imaging findings of thoracic ECD on cardiac MRI, chest CT and fludeoxyglucose positron emission tomography (FDG PET/CT) and provides a primer for radiologists seeking to familiarize themselves with the cardiothoracic manifestations of this rare disease.

Cardiac

Cardiac and vascular ECD lesions develop as a result of periadventitial infiltration of the pericardium, myocardium and coronary arteries and occur in 40% of cases.9 Published consensus guidelines for baseline evaluation of ECD now recommend cardiac MRI in all patients to identify involvement and delineate extent of disease.4 Clinically, cardiac involvement by ECD can result in a variety of presentations. Some patients remain asymptomatic, while others develop arrhythmias, valvular heart disease, ischemia, or cardiac failure. Hypertension is present in up to 66%, which may be due to renal artery infiltration causing stenosis.4,10

Pericardial involvement occurs in 13–24% of patients and can result in diffuse thickening and/or pericardial effusion 5,9 (Figure 1) which can rarely lead to cardiac tamponade.9,11 Pericardial calcification is rare, occurring in 4% of patients.5 Myocardial infiltration occurs in 25–31%, most commonly affecting the right atrium and right atrioventricular groove (Figure 2) with up to 30% developing a right atrial "pseudotumour."5,9,11 The typical appearance on cardiac MRI is of focal hypointense infiltration on T1 and balanced steady state free precession (b-SSFP) images, which demonstrates post-contrast enhancement (Figure 2). Patchy late gadolinium enhancement of the myocardium in ECD patients may also be seen, in a non-coronary distribution.4 Periarterial infiltration of the coronary arteries (Figure 3) has been described in up to one-third of ECD patients, most commonly affecting the right coronary artery 7 and may cause vascular stenosis and territorial ischemia.5,9

Figure 1.

Figure 1.

Axial b-SSFP (b-SSFP) pulse sequence (A) and gadolinium enhanced T1 weighted (B) cardiac MR images showing diffusely thickened pericardium, particularly along the left ventricle free wall demonstrating post-contrast enhancement (arrows). b-SSFP,balanced steady state free precession.

Figure 2.

Figure 2.

Axial oblique b-SSFP cardiac MR images showing focal hypointense thickening of the right atrial wall (A) and atrioventricular groove (B). b-SSFP, balanced steady state free precession.

Figure 3.

Figure 3.

Select axial oblique b-SSFP pulse sequence cardiac MR images (A, B) exhibiting hypointense soft-tissue infiltration encasing the right coronary artery. b-SSFP, balanced steady state free precession.

On PET/CT, involvement of the heart and pericardium in patients with ECD may demonstrate abnormal uptake of FDG demonstrating a maximum standardized uptake value (SUVmax) ranging from 3.6 to 8.1.12 Detecting pathological FDG uptake in the heart can be challenging because physiologic cardiac FDG uptake is found to variable degree in predominantly the left ventricular wall. Physiologic cardiac uptake typically demonstrates a homogenous uptake pattern with smooth margins, while ECD often demonstrates a patchy heterogeneous uptake pattern, most commonly in atrial walls or along the interatrial septum (Figure 4). To reduce physiologic cardiac uptake, a high fat, low carbohydrate diet can be prescribed before PET imaging.13

Figure 4.

Figure 4.

Axial contrast-enhanced CT (A) and fused axial PET/CT (B) images demonstrating right atrial "pseudotumour" showing patchy heterogenous FDG uptake (arrows) in a patient with biopsy-proven ECD. ECD, Erdheim-Chester disease; FDG, fludeoxyglucose; PET, positron emissiontomography.

Mediastinal

ECD involves the mediastinum in up to 62% of patients.7 The most common manifestation is "sheathing" or infiltration of the thoracic aorta,5,7,12 which most frequently affects the arch (38%) (Figure 5) but can also involve the ascending and descending aorta (30%) (Figure 6). On MRI, periaortic ECD lesions are usually circumferential, hypointense on both T1 and b-SSFP pulse sequences, and demonstrate post-gadolinium enhancement. Perivascular mediastinal ECD lesions demonstrate moderate FDG uptake on 18F-FDG PET/CT with an average SUVmax of 4.5.12 Vascular sheathing less commonly affects the pulmonary arteries (Figure 7), which occurs in 2–8% of ECD patients.7 Infiltration of the mediastinal fat by ECD (Figure 8) is observed in up to 11%.7 Mediastinal lymphadenopathy (Figure 9) is uncommon, affecting 6–11% of patients and may demonstrate variable FDG uptake.7,12

Figure 5.

Figure 5.

Axial oblique b-SSFP pulse sequence MRI (A) and contrast-enhanced CT (B) and FDG PET/CT (C) images showing vascular "sheathing" of the aortic arch, which demonstrates mild FDG uptake with an SUVmax of 2.8. b-SSFP, balanced steady state free precession; ECD, Erdheim-Chester disease; FDG, fludeoxyglucose; PET, positron emission tomography; SUV max, maximum standardized uptake value.

Figure 6.

Figure 6.

Axial contrast-enhanced CT showing soft tissue "sheathing" of the descending aorta (A) demonstrating moderately intense FDG uptake on FDG PET/CT (B). FDG, fludeoxyglucose; PET, positron emission tomography.

Figure 7.

Figure 7.

Axial oblique b-SSFP pulse sequence (A) and gadolinium-enhanced T1 (B) cardiac MR images showing hypointense soft tissue infiltration of the pulmonary artery trunk demonstrating moderate post-contrast enhancement (arrows). b-SSFP, balanced steady state free precession.

Figure 8.

Figure 8.

Contrast-enhanced axial CT (A) and axial PET/CT (B) images showing soft tissue infiltration of the posterior mediastinum (arrow) demonstrating FDG uptake with SUVmax 4.6. FDG, fludeoxyglucose; PET, positron emission tomography; SUV max, maximumstandardized uptake value.

Figure 9.

Figure 9.

Axial non-contrast CT (A) and corresponding FDG PET/CT (B) exhibiting upper right paratracheal lymphadenopathy demonstrating intense FDG uptake with SUVmax 7.4. FDG, fludeoxyglucose; PET, positron emission tomography; SUV max, maximumstandardized uptake value.

Pleural

Involvement of the pleura by ECD occurs in 15–41% of patients.6,12 Pleural ECD lesions may result in focal or diffuse pleural thickening seven that can demonstrate variable FDG uptake12 (Figure 10). Pleural effusions can be unilateral or bilateral and are encountered over three times more frequently than pleural thickening.6 Pleural effusions and pleural thickening co-exist in up to 7% of patients.7

Figure 10.

Figure 10.

Axial contrast-enhanced CT (A) and FDG PET/CT showing thickened left lower lobe posterior pleura with mild uptake of FDG (SUVmax 2.1). FDG, fludeoxyglucose; PET, positron emission tomography; SUV max, maximumstandardized uptake value.

Lung parenchymal

Lung parenchymal involvement is reported in up to two thirds of patients with ECD.1,6 Pulmonary ECD is challenging to distinguish from other causes of interstitial lung disease (ILD). Therefore, radiological evidence of other sites of systemic ECD involvement, in addition to tissue sampling may be required to ascertain a definitive diagnosis.5,6 Parenchymal disease occurs due to the spread of disease by infiltrating histiocytes and can result in peribronchovascular, interlobular septal and fissural involvement. Histopathological evaluation demonstrates diffuse interstitial histiocytes with foamy cytoplasm embedded in fibrosis.5 Many patients remain asymptomatic. Symptoms can occur in 35% of patients with radiologically assessed “minimal” ECD related ILD and in 67% of those with moderate to severe ILD, most commonly dyspnea and chronic cough.5 Pulmonary fibrosis/honeycombing is rare.5

The most commonly described CT finding in pulmonary ECD is interlobular septal thickening (Figure 11), which occurs in approximately two-thirds of patients can be focal or diffuse, and is most frequently smooth in appearance. Focal involvement is more common.5,6 Septal thickening can have a variable distribution but when diffuse septal thickening is present it more commonly demonstrates a peripheral and predominantly lower lobe distribution.7 Pulmonary nodules (Figure 12) are encountered in between 21–62% of patients. The distribution of nodules can be variable and both centrilobular nodularity, and subpleural/perifissural nodularity has been reported. Nodules tend to have an upper lobe predominance.2,7,8 Ground glass opacities are seen in 36–42% of patients6,7 and may exhibit a subpleural and/or peribronchovascular distribution.7 Areas of bronchial wall thickening 7 and consolidation are rare.6 Lung parenchymal involvement by ECD can demonstrate variable FDG avidity with mean SUVmax values of 4.3.12 Tissue sampling of pulmonary nodules and ground glass opacities may be required to establish a diagnosis of ECD.

Figure 11.

Figure 11.

Selected lung windowed images from non-contrast CT showing diffuse, bilateral interlobular septal thickening, fissural thickening (B) and small left pleural effusion in a patient with biopsy proven ECD. ECD, Erdheim-Chester disease.

Figure 12.

Figure 12.

Axial lung window chest CT (A) and FDG PET/CT (B) images showing an FDG-avid subpleural right upper lobe pulmonary nodule. Histopathology demonstrated histiocytes and foamy macrophages consistent with ECD. ECD, Erdheim-Chester disease; FDG, fludeoxyglucose; PET, positron emissiontomography.

Diagnosis and response assessment

Given the rarity, complexity and multisystemic clinical presentation of ECD, close multidisciplinary cooperation and collaboration is required to make a definitive diagnosis of cardiothoracic ECD and plan treatment.4 If ECD is included in the differential diagnosis based on characteristic cardiothoracic involvement additional imaging suggested by the radiologist may be useful in assessing for other involved sites, in particular the lower limbs/femora, the retroperitoneum and the brain. Once a pathological diagnosis has been established, concensus guidelines for the initial assessment of patients recommend CT of the chest/abdomen/pelvis, whole body PET/CT, Brain MRI and cardiac MRI. Historically, 3–5 year survival rates for patients with ECD have ranged from 43 to 68%.1,2,14 Baseline imaging to define the extent of disease, biopsy confirming characteristic histopathology, and analysis to assess BRAFV600E mutational status are vital prior to commencing therapy.

Promising results for select ECD patients with BRAF mutations have recently been reported with the BRAFV600E inhibitor vemurafinib.14,15 Between 38 and 68% of patients with ECD harbour a BRAFV600E mutation, which activates the RAF-MEK-extracellular signal-regulated kinase signalling pathway.4,10 PET/CT may have a role in assessing treatment response in this context (Figure 13). In a recently published efficacy and safety analysis of 22 adult patients with BRAF V600E mutant ECD treated with vemurafinib, response was analyzed using both Response Evaluation Criteria in Solid Tumours 1.1 (RECIST1.1) and modified PET Response Criteria in Solid Tumours (PERCIST) in a subset of patients. The overall response rate (ORR) as assessed by RECIST 1.1 was 62%, while the ORR as assessed by PERCIST was 100% (including 80% who had a complete metabolic response), suggesting that RECIST based assessment may underestimate benefit in this context.15

Figure 13.

Figure 13.

Fused PET/CT (A), contrast enhanced CT (B) and PET images (C) demonstrating patchy uptake in myocardium in a patient with cardiac ECD prior to treatment with BRAFV600E inhibitor with subsequent resolution of abnormal uptake observed on interval follow-up imaging at 3 months (D, E, F). ECD, Erdheim-Chester disease; PET, positron emissiontomography.

Established medical therapies for ECD patients without a BRAF mutation include interferon-α and cladribine. Surgery and radiotherapy have a limited role in the treatment of thoracic ECD lesions, but may be utilized for the treatment of localized mechanical complications, for example surgical repair of cardiac valvular dysfunction.

Conclusion

ECD is a rare multisystem disease that frequently involves cardiothoracic structures. Radiologists play a key role in the management of this complex patient cohort. Knowledge of the imaging findings of cardiothoracic ECD can facilitate early diagnosis, allow more prompt multidisciplinary input and enable timely treatment with novel agents where possible, potentially reducing morbidity and mortality.

Contributor Information

Jeeban Paul Das, Email: jeeban.paul.das@gmail.com.

Lola Xie, Email: xiel@mskcc.org.

Chris C Riedl, Email: riedlc@mskcc.org.

Sara A Hayes, Email: hayess@mskcc.org.

Michelle S Ginsberg, Email: ginsberm@mskcc.org.

Darragh F Halpenny, Email: halpennd@mskcc.org.

REFERENCES

  • 1.Veyssier-Belot C, Cacoub P, Caparros-Lefebvre D, Wechsler J, Brun B, Remy M, et al. Erdheim-Chester disease clinical and radiologic characteristics of 59 cases. Medicine 1996; 75: 157–69. doi: 10.1097/00005792-199605000-00005 [DOI] [PubMed] [Google Scholar]
  • 2.Arnaud L, Pierre I, Beigelman-Aubry C, Capron F, Brun A-L, Rigolet A, et al. Pulmonary involvement in Erdheim-Chester disease: a single-center study of thirty-four patients and a review of the literature. Arthritis & Rheumatism 2010; 62: 3504–12. doi: 10.1002/art.27672 [DOI] [PubMed] [Google Scholar]
  • 3.Ozkaya N, Rosenblum MK, Durham BH, Pichardo JD, Abdel-Wahab O, Hameed MR, et al. The histopathology of Erdheim–Chester disease: a comprehensive review of a molecularly characterized cohort. Mod Pathol 2018; 31: 581–97. doi: 10.1038/modpathol.2017.160 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Diamond EL, Dagna L, Hyman DM, Cavalli G, Janku F, Estrada-Veras J, et al. Consensus guidelines for the diagnosis and clinical management of Erdheim-Chester disease. Blood 2014; 124: 483–92. doi: 10.1182/blood-2014-03-561381 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Ghotra AS, Thompson K, Lopez‐Mattei J, Bawa D, Hernandez R, Banchs J, et al. Cardiovascular manifestations of Erdheim–Chester disease. Echocardiography 2019; 36: 229–36. doi: 10.1111/echo.14231 [DOI] [PubMed] [Google Scholar]
  • 6.Haroutunian S, O’Brien K, Estrada-Veras J, Yao J, Boyd L, Mathur K, et al. Clinical and histopathologic features of interstitial lung disease in Erdheim–Chester disease. JCM 2018; 7: 243.. pii: E243. doi: 10.3390/jcm7090243 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Mirmomen SM, Sirajuddin A, Nikpanah M, et al. Thoracic involvement in Erdheim-Chester disease: computed tomography imaging findings and their association with the BRAFV600E mutation. Eur Radiol 2018;. [DOI] [PubMed] [Google Scholar]
  • 8.Wittenberg KH, Swensen SJ, Myers JL. Pulmonary involvement with Erdheim-Chester disease: radiographic and CT findings. AJR Am J Roentgenol 2000; 174: 1327–31. [DOI] [PubMed] [Google Scholar]
  • 9.Haroche J, Amoura Z, Dion E, Wechsler B, et al. Cardiovascular involvement, an overlooked feature of Erdheim-Chester disease: report of 6 new cases and a literature review. Medicine 2004; 83: 371–92. [DOI] [PubMed] [Google Scholar]
  • 10.Haroche J, Amoura Z, Trad SG, Wechsler B, Cluzel P, Grenier PA, et al. Variability in the efficacy of interferon-α in Erdheim-Chester disease by patient and site of involvement: results in eight patients. Arthritis Rheum 2006; 54: 3330–6. doi: 10.1002/art.22165 [DOI] [PubMed] [Google Scholar]
  • 11.Haroche J, Cluzel P, Toledano D, et al. Images in cardiovascular medicine. cardiac involvement in Erdheim-Chester disease: magnetic resonance and computed tomographic scan imaging in a monocentric series of 37 patients. Circulation 2009; 119: e597–8. [DOI] [PubMed] [Google Scholar]
  • 12.Young JR, Johnson GB, Murphy RC, Go RS, Broski SM, et al. 18 F-FDG PET/CT in Erdheim–Chester Disease: Imaging Findings and Potential BRAF Mutation Biomarker. J Nucl Med 2018; 59: 774–9. doi: 10.2967/jnumed.117.200741 [DOI] [PubMed] [Google Scholar]
  • 13.Williams G, Kolodny GM. Suppression of Myocardial 18 F-FDG Uptake by Preparing Patients with a High-Fat, Low-Carbohydrate Diet. American Journal of Roentgenology 2008; 190: W151–W156PubMed PMID. doi: 10.2214/AJR.07.2409 [DOI] [PubMed] [Google Scholar]
  • 14.Hyman DM, Puzanov I, Subbiah V, Faris JE, Chau I, Blay J-Y, et al. Vemurafenib in Multiple Nonmelanoma Cancers with BRAF V600 Mutations. N Engl J Med 2015; 373: 726–36. doi: 10.1056/NEJMoa1502309 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Diamond EL, Subbiah V, Lockhart AC, et al. Vemurafenib for BRAF V600-mutant Erdheim-Chester disease and Langerhans cell histiocytosis: analysis of data from the Histology-Independent, phase 2, open-label VE-BASKET study. JAMA Oncol 2018; 4: 384–8. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from The British Journal of Radiology are provided here courtesy of Oxford University Press

RESOURCES