Beyond Pneumonia: A Case of Steroid-Responsive Pulmonary Eosinophilia With Atypical p-ANCA Positivity.

Persistent pulmonary opacities despite antimicrobial therapy should prompt consideration of diagnoses beyond infection. We present a 76-year-old man with heart failure with moderately reduced ejection fraction, coronary artery disease, type 2 diabetes mellitus, and eczema treated with dupilumab who presented with progressive dyspnea, dry cough, weakness, hypoxemia, and a recent episode of hemoptysis. He had recently been hospitalized for presumed pneumonia and underwent bronchoscopy at an outside hospital before being discharged on amoxicillin, without significant improvement. Chest computed tomography (CT) demonstrated diffuse bilateral ground-glass opacities (GGOs), and he was again treated empirically for multifocal pneumonia. His symptoms and oxygen requirement persisted despite antibiotics, and an extensive infectious evaluation was unrevealing. Repeat CT showed persistent bilateral GGOs with increasing left basilar consolidation. Bronchoscopy with bronchoalveolar lavage (BAL) demonstrated mixed inflammation with 13% eosinophils, and cytology showed acute inflammation with eosinophils. Although several organisms were detected in BAL specimens, Infectious Disease considered these findings to represent colonization, contamination, or viral reactivation rather than active infection. Peripheral eosinophilia was present, with a peak absolute eosinophil count of 1.695 × 10³/µL and a peak eosinophil percentage of 15%. Indirect immunofluorescence demonstrated repeatedly positive atypical perinuclear antineutrophil cytoplasmic antibody (p-ANCA), with titers increasing from 1:40 to 1:80, while myeloperoxidase (MPO) and proteinase 3 (PR3) antibodies remained negative. Organizing pneumonia, eosinophilic lung disease, medication-associated pulmonary inflammation, and ANCA-associated vasculitis, including eosinophilic granulomatosis with polyangiitis, were considered. Prednisone 60 mg orally daily was initiated for the suspected inflammatory pulmonary process. The absolute eosinophil count decreased from 0.79 × 10³/µL (10%) immediately before corticosteroid initiation to 0.00 × 10³/µL (0%) the following day, and the patient was documented on room air approximately 48 hours after corticosteroid initiation. However, concurrent heart-failure-directed therapy and possible cardiogenic pulmonary edema limited attribution of the improvement solely to corticosteroids. This case illustrates the diagnostic challenge of non-resolving pulmonary infiltrates when eosinophilic inflammation, abnormal but non-specific ANCA testing, medication exposure, infection, and cardiogenic pulmonary edema provide competing explanations.