Clinico-radiological correlation and prognostic value of baseline chest computed tomography in eosinophilic granulomatosis with polyangiitis

Abstract

Objectives

While chest high-resolution CT (HRCT) is correlated to severity and prognosis in asthma, it has not been studied in eosinophilic granulomatosis with polyangiitis (EGPA). Our objective is to study the prognostic value of baseline HRCT in EGPA patients.

Methods

Retrospective, multicentre observational study in three French hospitals, including EGPA patients with available chest HRCT before any systemic treatment. Two experienced radiologists blinded to clinical data evaluated HRCT images using semi-quantitative scoring. HRCT characteristics were correlated with clinical features and outcome.

Results

Among 46 patients, 38 (82.6%) had abnormal parenchymal findings on HRCT, including bronchial wall thickening (69.6%), mosaic perfusion (63.0%), ground-glass opacities (32.6%), bronchiectasis (30.4%), mucous plugging (21.7%) and consolidations (17.4%). Patients were clustered into three groups depending on HRCT features: ground-glass pattern, i.e. with ground-glass opacities with or without bronchial abnormalities (group 1, 28.3%), bronchial pattern (group 2, 41.3%) and extra-pulmonary pattern with no significant abnormality (group 3, 30.4%). Group 2 showed less frequent cardiac involvement (31.6 vs 46.2 and 42.9% in groups 1 and 3), more frequent positive ANCA (52.6 vs 0.0 and 14.3%) and higher eosinophil count (median 7510 vs 4000 and 4250/mm³). Group 1 showed worse prognosis with more frequent steroid-dependency (58.3 vs 11.1 and 28.6%) and requirement for mepolizumab (25.0 vs 11.1 and 7.1%). Conversely, group 2 showed a better outcome with higher rates of remission (88.9 vs 41.6 and 71.4%).

Conclusion

Chest HRCT at diagnosis of EGPA may have prognostic value and help clinicians better manage these patients.

Introduction

Eosinophilic granulomatosis with polyangiitis (EGPA; previously called Churg–Strauss syndrome) is a rare inflammatory disease characterized by eosinophil-rich small-vessel vasculitis, blood and tissue eosinophilia and severe asthma [1]. Eosinophilic granulomatosis with polyangiitis is less common than other ANCA-associated vasculitides, with an incidence estimated from 1 to 11 cases per million [2–4], which seems stable over time [5]. Importantly, this incidence rate is 3–6 times higher in asthma patients than in non-asthma patients [6, 7]. Asthma is present in >90% of patients with EGPA [8, 9] and is considered as a mandatory criteria in recent diagnostic criteria used in large randomized controlled trials [10]. Asthma is usually severe [11–13] and can precede vasculitis features by 8–12 years [9, 11]. Respiratory manifestations are of major concern in EGPA, as chronic asthma is the most frequent sequelae and pulmonary symptoms the main features at relapse [14], requiring an increase in glucocorticoids (GCs) dose and/or initiation of immunosuppressive or immunomodulatory agents.

High-resolution CT (HRCT) now has an established role in allergic asthma, as it successfully evaluates bronchial wall thickening, air trapping and bronchiectasis that are known to be correlated to disease severity and duration in many studies [15–23]. Moreover, HCRT is also correlated to treatment response in allergic asthma [24, 25]. Because of the rarity of EGPA, impact of HRCT in EGPA patients remains unknown, and even if few studies described CT findings in EGPA [26–31], none of them evaluated its correlation with clinical features and outcome. Here, we describe initial HRCT findings in 46 EGPA patients using a centralized review blind for clinical data, and we analyse its correlations with clinical features and patients’ outcomes.

Patients and methods

Study design and patients

We retrospectively included patients referred to three French departments of Internal Medicine or Pulmonology between 2006 and 2019 with a diagnosis of EGPA and an available chest HRCT at diagnosis, before any systemic treatment. Patients with previous treatment, either systemic CS or immunosuppressive therapy, were excluded from our study. All patients fulfilled either the ACR 1990 classification criteria [32] or the MIRRA criteria for diagnosis of EGPA [10]. Diagnosis of EGPA was reviewed by the investigators (F.D. and B. Terrier) by taking into account the entire medical history of participants. Patients considered to have alternative diagnosis of eosinophilic lung disorders were excluded. This study was conducted in compliance with the Good Clinical Practice protocol and the Declaration of Helsinki principles, and was approved by the local ethics committee from Cochin Hospital, Paris (approval number AAA-2021-08023).

Baseline measurements and assessments

Medical records were reviewed for demographic characteristics, clinical and biological findings. ANCA status was considered to be positive when ANCA were detected by ELISA either at diagnosis or later during the disease course. Data regarding treatments, especially GCs and inhaled CS (ICS) dose and the use of immunosuppressive or immunomodulatory agents were collected.

High-resolution CT

Two experienced chest radiologists (A.-L.B., M.-P.R.) blinded to clinical and outcome data reviewed HRCT images together, and made decisions concerning HRCT findings by consensus. Six relevant radiological items were retained and quantified using a semi-quantitative scoring (Supplementary Table S1, available at Rheumatology online) ranging from 0 (absent feature) to 3 (severe). These items included mucous plugging, bronchiectasis, bronchial wall thickening, mosaic perfusion, ground-glass opacities and consolidation. All HRCT parameters were defined according to the Fleischner Society criteria, and items were scored as in previous studies using semi-quantitative method [33, 34]. Mucous plugging and bronchiectasis were scored according to their extension, as all the bronchiectatic changes were mild and cylindrical when present. Bronchial wall thickening was assessed comparing the bronchial lumen diameter and external bronchial diameter. As ground glass opacities (GGO) can be caused by alveolar filling, interstitial thickening and/or increased capillary blood volume, the radiologists analysed the ‘intensity’, distribution and margins of GGO, as well as the size of pulmonary vessels in the lung parenchyma to minimize overlap between GGO and mosaicism. GGO typically appears as dense, ill-defined opacities often associated with consolidation and an underlying preserved and homogeneous perfusion. Mosaicism was considered in the presence of hypoattenuating geographic areas with small and rare vascular markings, associated with areas of relative hyperattenuation due to vascular redistribution in the spared lung. As most of the patients with EGPA suffer from asthma, and in the absence of underlying features suggesting primary vascular disease (evidence of pulmonary hypertension or cor pulmonale), mosaicism was considered to represent small airway disease even in the absence of expiratory slices. Bronchial score was then defined by adding mean value of bronchial wall thickening and mosaic perfusion scores.

Clustering

Based on the radiological features, patients were clustered into three groups depending on the predominant HRCT features: ground-glass pattern, i.e. with a ground-glass score ≥1 with or without signs of bronchial involvement (group 1); bronchial pattern, i.e. with a bronchial score ≥1 and ground-glass score <1 (group 2); and extra-pulmonary pattern, i.e. with both bronchial and ground-glass opacities scores <1 (group 3).

Outcome measures

Outcomes were defined according to disease status at 1 year of follow-up. Complete remission was defined as a BVAS, version 3, of 0 and the receipt of prednisolone or prednisone at a dose of ≤4.0 mg per day [35]. Partial remission was defined by a BVAS of 0 and the receipt of prednisolone or prednisone at a dose between 5 and 10 mg per day. Treatment failure was defined by inability to achieve disease remission with GCs dose <10 mg/day. Vasculitis relapses were recorded separately and were defined as the recurrence or worsening of EGPA-related manifestations, excluding asthma and/or ENT manifestations, following a period of remission ≥3 months, or the need for change or initiation of immunosuppressive or immunomodulatory agents and/or increase in prednisone dosage.

Statistical analysis

All results are expressed as median and interquartile range (IQR) for continuous variables and number (percentage) for categorical variables. Percentages were calculated on available data.

The study is reported according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement [36] for reporting of observational studies.

Results

Clinical features of EGPA patients

Diagnosis and clinical features of patients are summarized in Table 1. Among the 46 patients included, 24 patients were men (52.2%) with a median age of 52 (IQR 39–64) years. All patients had asthma (n = 46, 100.0%). Thirty-one of 37 (83.8%) were treated with ICS, and among them 26 patients (70.3%) received high dose ICS. Extra-pulmonary clinical manifestations included ENT involvement in 34 patients (73.9%), mostly chronic sinusitis (n = 24, 52.2%) and nasal polyposis (n = 12, 26.1%), cardiac involvement in 18 (39.1%), mostly myocarditis (n = 12, 26.1%) and myopericarditis (n = 6, 13.0%), neurological involvement in 17 (37.0%), mostly multiple mononeuropathy (n = 14, 30.4%), cutaneous symptoms in 14 (30.4%), mostly purpura (n = 10, 21.7%), and myalgia, gastrointestinal involvement and renal involvement in three cases (6.5%) each. ANCA were positive in 12 patients (26.1%), with MPO specificity in 8. Median peak eosinophil count at diagnosis was 5500 (2600–11 300)/mm³, with 30.4% of patients having eosinophil count >10 000/mm³.

HRCT features and clustering at diagnosis

Imaging features from chest HRCT are shown in Fig. 1 and Table 2. Of the 46 patients, 38 (82.6%) had pulmonary abnormalities on baseline HRCT, including 32 (69.6%) with significant abnormalities (i.e. scoring ≥1). Main radiological features were bronchial wall thickening (n = 32, 69.6%), mosaic perfusion (n = 29, 63.0%) and GGO (n = 15, 32.6%). Other HRCT findings included mild-to-moderate bronchiectasis (n = 14, 30.4%), mucous plugging (n = 10, 21.7%) and consolidations (n = 8, 17.4%).

Patients were clustered into three groups depending on the predominant HRCT features (Figs 1 and 2). Group 1 included patients with a ground-glass pattern with or without signs of bronchial involvement (n = 13, 28.3%) (Fig. 3). Group 2 included patients with bronchial pattern without GGO (n = 19, 41.3%) (Fig. 4). Group 3 included patients with an extra-pulmonary pattern, i.e. without any major chest HRCT abnormality (n = 14, 30.4%).

Patients with bronchial pattern were older [median age 65 (52–72) vs 48 (39–54) and 47 (35–53) years in groups 1 and 3, respectively], had less frequent cardiac involvement (31.6 vs 46.2 and 42.9%), more frequent neurological involvement (47.4 vs 30.8% and 28.6%) and positive ANCA (52.6 vs 0.0 and 14.3%), and higher blood eosinophil count [median 7510 (2275–12 875) vs 4000 (2750–7600) and 4250 (2100–12 983)/mm³]. Eosinophilic alveolitis defined with a cut-off of 25% of eosinophils in bronchoalveolar lavage was evidenced in nine patients (19.6%), five of them with ground-glass pattern and four with extra-pulmonary pattern. Median eosinophil count in bronchoalveolar lavage fluid was 59.9% (40.0–78.8) for a total cellular count of 300 000 (160 000–320 000) cells/ml. Patients with ground-glass pattern had more frequent ENT manifestations (92.3 vs 68.4 and 64.3% in groups 2 and 3, respectively).

Outcomes

At 1 year of follow-up, 23/44 patients (52.3%) of patients achieved complete remission, 8/44 (18.2%) partial remission, 13/44 (29.5%) had treatment failure and 10/44 (22.7%) presented at least one relapse.

Patients from group 1 had a worse outcome with more frequent treatment failure (58.4 vs 11.1 and 28.6% in groups 2 and 3, respectively). Also, they were more frequently treated with mepolizumab (25.0 vs 11.1 and 7.1% in groups 2 and 3, respectively).

In contrast, patients with bronchial pattern had a better outcome with less frequent treatment failure, less frequent relapses (11.1 vs 25.0 and 35.7% in groups 1 and 3, respectively) and higher rates of complete remission at 1 year (77.8 vs 33.3 and 35.7%).

Discussion

We report here a retrospective observational study of baseline HRCT findings in EGPA and their correlations with disease presentation and outcome. HRCT was useful to classify EGPA patients according to clinical presentation and evolution, according to the presence of GGO and bronchial involvement that delineates three groups. Patients with bronchial pattern on HRCT were older, had less frequent cardiac involvement, more frequent neurological involvement and positive ANCA, and higher blood eosinophil count, and a better outcome (complete remission achievement and less relapse). In contrast, patients with predominance of GGO had more frequent ENT manifestations and a worse outcome with more frequent treatment failure and more frequent disease requiring mepolizumab.

Very few studies have described HRCT findings in EGPA. As in our study, main HRCT findings described were bronchial wall thickening, GGO, but also consolidation and bronchiolar nodules [26–31]. Nakamoto et al. reported that HRCT findings were correlated to severity and duration of asthma in EGPA [29] and they found a higher incidence of bronchiectasis in patients with severe asthma. They also detected higher rates of GGO and bronchial wall thickening in patients with a duration of asthma of 5 years or more. However, while they studied asthma severity at time of HRCT, whether HRCT findings were correlated with extra-pulmonary manifestations of EGPA and outcome was not studied. Kim et al. proposed a distinction of pulmonary features between airway pattern (with bronchial wall thickening, bronchial dilation, tree-in-bud sign or mosaic perfusion as predominant abnormalities) and airspace pattern (with parenchymal opacities such as GGO and consolidation as predominant feature) in 19 EGPA patients, and this classification was related to underlying histological features [37]. In this small study, patients with airspace pattern responded more readily to therapy than those with airway pattern, while we observed that patients with GGO showed a worse outcome. This discrepancy could be related to different definitions of groups, different study populations with potential other confounding factors, and the small sample size of both studies.

Importantly, we found that bronchial pattern was associated with less frequent cardiac involvement and more frequent positive ANCA, contrasting with more frequent cardiac involvement and less frequent positive ANCA in patients with ground-glass pattern (group 1) and the extra-pulmonary group (group 3). These findings are in agreement with those previously reported, showing that negative ANCA status was associated with cardiac involvement [9, 38–40]. This could suggest a common pathophysiology between pulmonary ground-grass infiltration and cardiac involvement, maybe mostly driven by eosinophils in tissues (‘eosinophilic phenotype’), and opposed to bronchial and systemic vasculitis involvement (‘vasculitis phenotype’) which could be more ANCA-driven [41]. More studies focusing on bronchial biopsies or bronchoalveolar lavage findings could help to a better comprehension of this clustering. This is of importance as EGPA is one of the most common systemic disease affecting the heart [42], and cardiac involvement was shown to be a major contributor of EGPA-related death [9, 38, 43].

Patients with ground-glass pattern had lower blood eosinophil count than patients with bronchial pattern, in contrast to half of patients with GGO showing eosinophilic alveolitis compared with none in bronchial pattern. These findings suggest that patients with GGO within the ground-glass pattern had features of higher extravascular tissue infiltration by eosinophils despite lower peripheral eosinophil count, as supported by more frequent eosinophilic alveolitis and cardiomyopathy. Inconsistent correlations between blood and tissue eosinophilia and organ damage have been largely reported in the literature. However, the duration and severity of eosinophilia which could lead to tissue damage remains unknown, and such consequences are probably more related to intrinsic features of eosinophils [44, 45].

In our study, 26.1% of patients had positive ANCA, which is slightly lower than previous cohort studies in which ANCA were reported in 30–40% of patients [9]. This very small difference could be related to our inclusion criteria which included a chest CT scan available at diagnosis and before any systemic treatment, therefore increasing potentially the proportion of patients with an ‘eosinophilic’ phenotype, including respiratory manifestations, more frequent in ANCA-negative patients [41].

Our study has some limitations, especially its retrospective design and the small number of patients. However, HRCT evaluation was centralized and performed blinded to clinical data by the expert chest radiologists. Also, improvement in image processing over time has made it possible to quantitatively assess the bronchial tree and lung parenchyma, allowing a detailed analysis of large and small airways and their remodelling in asthma [46, 47], which could be also of interest in EGPA. Next, we chose to exclude patients that had been previously treated to avoid bias in HRCT findings, which drastically reduced the number of analysed patients. Further studies with more patients are needed to confirm our findings on one hand, and validate chest HRCT semi-quantitative scoring in EGPA patients in the other hand. A second limitation is that patients with extra-pulmonary pattern are a heterogeneous group regarding outcome, with more frequent relapses (35.7% vs 25.0% and 11.1% in groups 1 and 2, respectively) but also high rates of remission (complete or partial) (71.4% vs 41.6% and 88.9% in groups 1 and 2, respectively) preventing drawing conclusions about patients with no major HRCT abnormalities at baseline.

Overall, chest HRCT at the onset of pulmonary manifestations in EGPA diagnosis is correlated to clinical features and patients’ outcome, especially while studying GGO and bronchial involvement. Patients with bronchial-predominant pattern are more frequently ANCA positive, with higher eosinophilia and less frequent cardiac involvement and have higher rates of remission. In contrast, patients with ground-glass predominant pattern have a poorer outcome. As in allergic asthma, HRCT could help clinicians to more appropriately manage EGPA patients.

Supplementary material

Supplementary material is available at Rheumatology online.

Data availability

The data that support the findings of this study are available from the corresponding author, B.Terrier, upon reasonable request.

Contribution statement

F.D. collected data. F.D. and B.Terrier analysed and interpretated clinical data. A.-L.B. and M.-P.R. reviewed HRCT images. B.Terrier designed and supervised the study. C.T., N.L., X.P., L.M. and L.G. provided and cared for study patients, and critical revision of the article. B. Thoreau performed statistical analyses.

Funding

No specific funding was received from any bodies in the public, commercial or not-for-profit sectors to carry out the work described in this article.

Disclosure statement: The authors have declared no conflicts of interest.

References