This is the story of a typical teenage boy with pneumonia that turned out to be anything but typical. He presented to our emergency department with a five-day history of malaise and a three-day history of cough, fever, and progressively worsening shortness of breath. His initial vital signs were a temperature of 38.7° C, with a heart rate of 110, blood pressure of 109/77 and an oxygen saturation of 90% on room air.
On physical exam the patient exhibited increased work of breathing with scattered wheezing and crackles throughout both lung fields. Tests revealed that his initial white blood cell count was 22,000 with 79% neutrophils, 10% eosinophils, 6% lymphocytes, and 5% monocytes, hemoglobin of 15 and platelets of 339. Electrolytes and liver function tests were normal.
A chest X-ray was performed and illustrated diffuse, bilateral nodular infiltrates. The patient received nebulized albuterol and atrovent without much improvement in his oxygen saturation or his work of breathing; therefore, he was placed on 2 liters of supplemental oxygen. He was started on ampicillin for a presumed diagnosis of community-acquired bacterial pneumonia and admitted to the intensive care unit.
Over the next several hours, his work of breathing progressively increased requiring escalation of his ventilatory support to high flow nasal canula, BiPap and ultimately intubation and mechanical ventilation. The patient’s antibiotics were broadened to vancomycin and cefotaxime. A CT scan of his lungs revealed peripherally located bilateral nodular and hazy ground-glass opacities. At this point, the infectious diseases team was consulted and I, as the fellow, first heard the story.
In the beginning, it seemed like a familiar story: a previously well teenager suddenly becomes ill with fever and severe respiratory distress, is found to have some unusual imaging findings, and we are consulted to help figure out why.
My mind quickly began considering the broad differential and battery of tests I have learned to send in patients like this one: a viral respiratory panel looking for things like influenza or adenovirus, testing for atypical bacteria such as Mycoplasma pneumoniae, Legionella pneumophila, and Chlamydophila pneumoniae, cultures and cytology of the upper respiratory tract obtained during bronchoscopy, and an HIV test to assess for risk of Pneumocystis jiroveci pneumonia.
In the meantime, the patient would continue to receive antibiotics while I gathered more information. In the end, I suspected, this patient would turn out like many others in that he would eventually get better, our evaluation would fail to identify a specific etiology, and we would likely attribute his illness to an unidentified virus or atypical bacterial infection.
This patient, however, took a significantly different course. In discussion with the pulmonologist and radiologist, the peripheral distribution of ground-glass opacities on chest CT was reconsidered and raised suspicion for eosinophilic pneumonia. This suspicion was confirmed with bronchoscopy, which revealed a WBC count of 200, with 3% neutrophils, 40% lymphocytes, 2% monocytes and 55% eosinophils.
The patient was started on methylprednisolone 60 mg every 6 hours and within 72 hours was extubated to room air. His fevers resolved within 24 hours of starting steroids. Given this dramatic improvement, antibiotics were stopped and he was discharged home on hospital day 10 to complete a 2 week course of steroids.
What I learned was this: idiopathic acute eosinophilic pneumonia is a relatively rare disease, usually occurring in young adults and more commonly in men than in women, but can nevertheless occur in any age group. Generally, patients are previously well without any specific history of atopy, though eosinophilic pneumonia is considered by some experts to be a hypersensitivity reaction to some unknown inhaled allergen.
This theory is supported by the observation that some patients developed eosinophilic pneumonia after smoking cigarettes for the first time, after new outdoor activities, or after exposure to a large burden of dust or smoke – interestingly, our patient had recently moved boxes out of a friend’s basement prior to the onset of his symptoms.
Like the patient in question, patients typically present with a febrile illness of relatively short duration (i.e., less than 7 days) with associated nonproductive cough, myalgias, malaise, chest pain, and shortness of breath. It is not unusual for this to progress to hypoxemic respiratory failure requiring mechanical ventilation as it did in our patient. With regard to laboratory studies, a neutrophilic predominant leukocytosis is the most common initial finding on CBC with increasing eosinophilia noted later.
The percent of eosinophils found on bronchoalveolar lavage, however, is typically greater than 25%. Imaging generally shows peripherally located ground-glass opacities, nodules and reticular infiltrates, as in our patient, often with pleural effusions. Treatment for acute eosinophilic pneumonia is with steroids given for a relatively short course of 2-4 weeks and, like our patient, patients respond rapidly and uniformly to this therapy.
What I also learned was that success in fellowship, like much of medical education, seems to at a very basic level rely on pattern recognition. In general, this process serves us well and is rewarded. Most of the time, a teenager who seems to have a viral or atypical pneumonia will have exactly that. But there are times when this process fails us, when our cognitive biases blind us to the true diagnosis, and this case underscores the importance of questioning our assumptions.