COP or Eosinophilic Pneumonia? When the Pathology Is Inconclusive, Treat Both

A 50-year-old African American male with hypertension and a 20 pack-year smoking history presented to the emergency department with two days of respiratory distress and hypoxia, after three weeks of progressive dyspnea. He had no history of COVID-19 and was unvaccinated. On examination he had diffuse bilateral crackles more prominent at the bases, with accessory muscle use. He was placed on 15 L/min supplemental oxygen.

He was started on BiPAP (IPAP 12, EPAP 6, FiO2 100%) but a repeat ABG confirmed ARDS (PaO2/FiO2 ratio = 76). After 24 hours he was transitioned to high-flow nasal cannula at 60L/100% FiO2. IV ceftriaxone and azithromycin were started empirically for sepsis with possible pneumonia, and he was admitted to the ICU. Despite continued antibiotics, his condition failed to improve, and repeat WBC normalized to 9.4 K/UL with normal procalcitonin — making bacterial infection increasingly unlikely.

On hospital day five, high-resolution CT confirmed bilateral patchy consolidations with peripheral predominance. Bronchoscopy with BAL showed 15% eosinophils — elevated but below the classic ≥25% threshold for idiopathic eosinophilic pneumonia (IEP). A transbronchial biopsy showed chronic inflammation without organizing features — the pathology was indeterminate. Given the clinical picture — rapidly progressive hypoxia, peripheral consolidations, peripheral eosinophilia, negative infectious and autoimmune workup — empirical high-dose methylprednisolone (60 mg IV q6h) was started for suspected COP or IEP.

Figure 1

HRCT Day 5 — Bilateral Peripheral Consolidations

Axial and coronal CT images demonstrating bilateral peripheral consolidations with adjacent faint ground-glass opacification. This pattern is nonspecific but characteristic of both COP and idiopathic eosinophilic pneumonia — both entities favor peripheral (subpleural) distribution and produce this radiographic appearance. In the context of ARDS, peripheral eosinophilia, and negative infectious workup, these imaging findings supported empirical high-dose steroid initiation without awaiting a definitive pathologic diagnosis.

Within days of steroid initiation, the patient showed dramatic clinical improvement. By hospital day eight he had been weaned to 10 L/min nasal cannula. A slow steroid taper over the subsequent week allowed continued improvement in oxygen requirements. He was discharged on hospital day 15 requiring 4 L/min of home oxygen. Due to loss to follow-up, the possibility of relapse — which would favor COP over IEP — could not be established.

Figure 2

Hospital Course Timeline

Timeline illustrating the trajectory of oxygen requirements, key diagnostic interventions, and treatment milestones across the 15-day hospitalization. The pivotal inflection point — empirical methylprednisolone initiation on day 5 following negative infectious workup and equivocal biopsy — produced a rapid, progressive improvement in oxygenation that continued through discharge. This steroid response was both therapeutically definitive and diagnostically confirmatory of an inflammatory ILD, even in the absence of pathologic certainty between COP and IEP.

Discussion

COP and idiopathic eosinophilic pneumonia (IEP) occupy overlapping clinical territory, and this case demonstrates what happens when the usual diagnostic tools fail to distinguish between them. Both present with rapidly progressive hypoxia, bilateral interstitial infiltrates with peripheral predominance, negative infectious workup, and dramatic steroid responsiveness. The distinguishing features are largely probabilistic rather than definitive at the individual patient level: COP tends toward lower zone and subpleural consolidation with mass-like lesions; IEP favors upper zone distribution with more ground-glass; BAL eosinophilia ≥25% strongly favors IEP while COP typically shows mononuclear predominance; peripheral blood eosinophilia is more prominent and persistent in IEP. In this patient, the blood eosinophilia favored IEP, but the BAL eosinophil count (15%) fell below threshold, the CT distribution was diffuse rather than upper-predominant, and the biopsy was indeterminate. The diagnosis remained genuinely uncertain.

The most important clinical lesson is that diagnostic uncertainty between COP and IEP should not delay steroid initiation. Both conditions are treated with systemic corticosteroids. Both carry the risk of relapse — though COP relapses more commonly during taper, while complete acute IEP typically resolves without relapse. The pragmatic approach in this case — empirical methylprednisolone on clinical grounds after exhausting the infectious differential — was correct and produced a life-saving response. Delaying treatment to achieve histologic certainty in a deteriorating patient with ARDS-level hypoxia is not justified when the clinical-imaging-laboratory constellation points toward an inflammatory ILD. The caveat is that biopsy of any discrete mass-like lesions should be obtained before steroids are started, as steroid treatment alters the pathologic picture; in this patient no mass lesions were present.

The failure of transbronchial biopsy to provide a specific diagnosis is also informative. Small biopsy samples from transbronchial forceps frequently yield indeterminate results even in pathologically definable entities — the organizing pneumonia pattern requires sufficient alveolar tissue to demonstrate the characteristic granulation tissue plugs, and a small sample from a non-representative area will simply show chronic inflammation. Surgical VATS biopsy would have provided more tissue from a targeted area, potentially reaching a more specific diagnosis, but given the patient's already fragile oxygenation, surgical risk was prohibitive. In less acute presentations where biopsy is warranted, VATS remains the gold standard for ILD histology.