in ltrate. In our EGPA series, clinical manifestations included hypertension, pericarditis (15.1%), valvopathy, eosinophilic myocarditis, and/or congestive heart failure (16.4%), and 31% of the deaths were attributed to cardiac involvement [23].

However, the real frequency of EGPA cardiac disease may be underestimated. Detailed cardiac workups of 32 ambulatory EGPA patients considered to be in remission revealed that 62% had cardiac involvement: among them, 60% had echocardiographic defects, 66% had electrocardiogram (EKG) abnormalities, and 62% had MRI anomalies, but only 26% of those patients were symptomatic [43].

Gastrointestinal Involvement

Gastrointestinal involvement, often severe, is another FFS poor prognosis factor [35, 42]. Clinical manifestations are often nonspeci c, with abdominal pain, nausea, vomiting, diarrhea, and/or intestinal hemorrhage. Bowel perforation, the most severe digestive event, is associated with high mortality [42]. Ischemic colitis or mesenteric ischemia may require surgical intervention [44]. When obtained, biopsies can show signs of vasculitis, extravascular granulomas, sometimes mimicking pseudopolypoid lesions, and eosinophilic in ltrates. Thickening of the esophageal wall and sometimes other parts of the gastrointestinal tract could be indicative of eosinophilic in ltration; it disappears under glucocorticoids.

Renal Manifestations

EGPA kidney manifestations are less common than in other AAVs, affecting 16–22% of EGPA patients. In our series, the mean ± SD creatinine level at diagnosis was 89.1 ± 38.6 μmol/L (1.01 ± 0.44 mg/dL), and 12.8% of the patients had abnormally elevated proteinuria [23]. Histology of EGPA patients’ renal biopsies usually nds crescentic glomerulonephritis, which can be focal or diffuse, and less frequent renal vasculitis, eosinophilic in ltrates, or granuloma(s) [45]. Glomerulonephritis in EGPA is usually associated with anti-MPO-ANCA positivity. Ureteral stenosis has been described but only rarely [46].

Ophthalmological Manifestations

Ophthalmological involvement, including uveitis, episcleritis, ischemic vasculitis, and/or orbital infammatory pseudotumor with conjunctival involvement, has been described [47].

Complementary Investigations

Blood eosinophilia is an EGPA constant; it is also an ACR classi cation criterion, with eosinophil counts exceeding 15,000/mm3 or 10% of total leukocytes as a diagnostic criterion. The mean ± SD eosinophil count for patients in our series was 7569 ± 6428/mm3 [23], but eosinophil counts can exceed 50,000/mm3 [48]. Eosinophilia may rapidly decline within a few days of starting EGPA treatment. Although isolated eosinophilia can be a marker of EGPA evolution, it is probably not enough to diagnose a relapse. During follow- ­up, eosinophil counts rise frequently without engendering clinical manifestations, usually during glucocorticoid tapering, so close monitoring is essential.

Infammatory syndrome is documented in ~80% of EGPA patients, with elevated C-reactive protein at 66.9 mg/L or a high erythrocyte sedimentation rate of 56% in our series [23].

About a third of EGPA patients are ANCA-positive, mostly with a perinuclear immunofuorescence labeling (P-ANCA) pattern, and, according to enzyme-linked immunosorbent assay (ELISA), ~70% of the ANCAs are MPO-­ speci c.Antinuclear antibodies are not found, but rheumatoid factor can be present [49].

For EGPA patients with alveolar hemorrhage, BAL fuid can contain red blood cells and the Golde score is elevated. In other contexts, BAL can contain infammatory cells, predominantly eosinophils [38].

When biopsies are obtained, histological examination has good sensitivity for con rmation of the EGPA diagnosis [34, 44]. Biopsies may contain three types of lesions, but rarely simultaneously: (1) necrotizing small-to-medium-sized vessel vasculitis, showing brinoid necrosis of the media and pleomorphic cellular in ltrate(s), mainly eosinophils; (2) typical but nonspeci c extravascular granuloma(s), with central necrosis and epithelioid cells, which can also be seen in other vasculitides (e.g., GPA) or autoimmune diseases; and

(3) eosinophil in ltration of arterial walls and adjacent tissues of any organ [50].

Chest radiographs can contribute to diagnosing EGPA and its lung manifestations. In a study on 91 patients, chest X-ray abnormalities were seen in 58% of the patients at diagnosis and included ground glass attenuation (29%), consolidation (29%), pleural effusion (10%), and/or nodules or mass(es) (9%). These anomalies were frequently bilateral (58%) [38]. Chest CT scans can evaluate lung abnormalities more precisely, as they can visualize ground glass opacities (39%), bronchial wall thickening (32%), and/or micronodules <3 mm (24%).

Echocardiography or cardiac MRI can identify cardiac involvement and detect abnormalities in 38% of patients who are asymptomatic or have EKG abnormalities [43]. However,