A 22-year-old woman with a history of bronchial asthma was transferred to our hospital because of severe dyspnea and hypereosinophilia. She had no history of smoking, regular medication use, or any relevant family history of thromboembolic disorders. Initially, she was admitted to another hospital with cough, fever, and
exertional dyspnea (day 1). Laboratory findings revealed elevated eosinophil counts and inflammatory marker levels. Chest computed tomography (CT) demonstrated bilateral ground-glass opacities and portal vein thrombosis (Fig. 1). Despite antibiotic therapy, her condition deteriorated. Bronchoscopy performed on day 4 revealed a marked eosinophilic predominance in bronchoalveolar lavage fluid. Differential cell counts showed eosinophils at 86% (reference range: <1%), neutrophils at 5% (reference range: <3%), lymphocytes at 6% (reference range: 10–15%), and macrophages at 3% (reference range: ≥85%). The total cell count was 5.77 × 10⁶ cells/mL (reference range: 0.7–2 × 10⁵ cells/mL). Although methylprednisolone pulse therapy was initiated for the suspected eosinophilic pneumonia, her respiratory status worsened, necessitating intubation and transfer to our hospital on day 5.
At the time of referral, her vital signs were as follows: body temperature, 37.1°C; blood pressure, 125/73 mmHg; pulse, 52 beats/min (regular); respiratory rate, 22 breaths/min; and oxygen saturation, 98% (mechanical ventilation in A/C mode; fraction of inspired oxygen, 0.6; pressure support, 12 cmH2O; positive end-expiratory pressure, 10 cmH2O). Physical examination revealed coarse breath sounds bilaterally, without wheezing and erythematous lesions on both lower extremities.
Laboratory examinations revealed the following results: white blood cell count, 124,000/µL (reference range: 3,300–9,400/µL); eosinophils, 10.6% (reference range: 0.0–7.0%); platelet count, 64,000/µL (reference range: 13,000–32,000/µL); C-reactive protein, 5.5 mg/dL (reference range: 0.0–0.2 mg/dL); and fibrin degradation products, 91 µg/mL (reference range: 0.0–5.0 µg/mL). Skin biopsy revealed eosinophilic infiltration in and around the blood vessels in the dermis and subcutaneous tissue. Serum IgE level was markedly elevated at 3,630 IU/mL (reference range: 0.0–173 IU/mL). Tests for the FIP1L1/PDGFRA fusion gene, antinuclear antibodies, myeloperoxidase anti-neutrophil cytoplasmic antibodies, proteinase 3 anti-neutrophil cytoplasmic antibodies, β-d-glucan, and cryptococcal antigen yielded negative results. Angiotensin-converting enzyme and IgG levels were within normal ranges. Serologic tests for parasitic infections were performed; all results were negative. With no evidence of secondary or neoplastic HES, a diagnosis of idiopathic HES was established.
The clinical course of the patient is shown in Fig. 2. Methylprednisolone pulse therapy and continuous intravenous infusion of heparin (22,000 units/day) were administered for 3 days. As the initial administration of heparin was complicated by endobronchial bleeding, the heparin dose was subsequently adjusted to maintain the activated partial thromboplastin time (aPTT) at approximately 37 s, corresponding to approximately 1.5 times the pre-heparin baseline, to minimize the risk of further bleeding. Following respiratory improvement and decreased eosinophil counts (230 cells/µL), mechanical ventilation was successfully discontinued on day 11. However, despite the ongoing steroid treatment and continuous oral rivaroxaban, eosinophil levels rose again (844 cells/µL), and D-dimer levels remained elevated (21.6 µg/mL, reference range: 0.0–1.0 µg/mL). Therefore, contrast-enhanced CT was performed on day 20, revealing that inferior vena cava thrombosis and pulmonary embolism had developed in addition to persisting portal vein thrombosis (Fig. 3), despite the improvement in pulmonary infiltrates (Fig. 4a). Given the progression of thrombosis and suspected involvement of eosinophils in thrombus formation, rapid elimination of eosinophils was deemed necessary. Therefore, the continuous intravenous infusion of heparin (22,000 units/day) was resumed, and benralizumab was administered on day 21 to target eosinophil-mediated thrombosis. However, because the aPTT was excessively prolonged to 80 s, the heparin dose was reduced to 15,000 units/day; thereafter, the aPTT was maintained at approximately 50 s. Subsequently, the blood eosinophil count decreased to 0 cells/µL, with a concurrent reduction in D-dimer levels (2.6 µg/mL). Abdominal ultrasonography on day 31 demonstrated resolution of portal vein thrombosis, and follow-up contrast-enhanced CT on day 40 revealed significant reduction in the size of multiple thrombi (Fig. 4b, c). Treatment was switched from heparin to oral apixaban, and the patient was discharged on day 42. The follow up over 6 months with gradual dose tapering of steroids and continuation of apixaban administration showed no recurrence.