Acute eosinophilic pneumonia after recent start of smoking.

In November, 2013, a 16-year-old girl was referred with 2 days of fever, chest pain, dry cough, and progressive dyspnoea. She had a history of psychotic disorder, anorexia, pervasive developmental disorder not otherwise specifi ed, and self-harm. She had been admitted to a psychiatric facility 2 months previously and had been taking aripirazole for more than 3 months. She had no history of drug or toxin abuse but 2 weeks earlier had started smoking 10 cigarettes a day. 24 hours before admission her family doctor had started amoxicillin for presumed community acquired pneumonia. On examination she was unwell, with dyspnoea, tachypnoea, and oxygen saturation of 90% breathing ambient air. Auscultation showed extended expiration with bilateral crackles. Chest radiograph showed bilateral infi ltrates (appendix) and her infl ammatory markers were raised. Nebulised ipratropiumbromide and salbutamol were ineff ective, so we started oral amoxicillin–clavulanic acid and azithromycin for suspected atypical pneumonia. On day 3 she deteriorated with persistent fever and progressive tachypnoea and hypoxia. Chest sounds were absent on the right with diff use crackles on the left side. Repeat chest radiograph showed progressive bilateral infi ltrations and a right-sided pleural eff usion, but unexpectedly C-reactive protein had decreased from 92 mg/L to 32 mg/L. Because our patient had a pneumonia that was not resolving with standard antibiotic treatment we widened our diff erential diagnosis to include antimicrobial failure due to resistant or unusual micro-organisms (such as legionella), infectious complications (such as Staphylococcus aureus empyema), and noninfectious causes such as chemical pneumonitis, eosinophilic pneumonia, vasculitis, or diff use lung disease. We changed the antibiotics to fl ucloxacillin and ciprofl oxacin. Tests for respiratory viruses, HIV, chlamydia, legionella, tuberculosis, Pneumocystis jirovecii, nocardia, aspergillus, and toxoplasma were negative. Blood cultures were negative. Autoantibody and complement screening were normal. Thoracic CT showed ground glass opacities, peripheral patchy nodular and confl uent consolidations, and bilateral pleural eff usions (appendix). On the basis of her clinical condition, CT scan fi ndings, and recent start of smoking, we suspected an acute eosinophilic pneumonia. Bronchoscopy and bronchoalveolar lavage (BAL) showed no abnormalities apart from eosinophilia (27%) in the BAL (appendix). At the same time peripheral eosinophilia increased from 1% to 18%. Eosinophilic lung diseases are rare. Primary eosinophilic lung diseases consist of simple pulmonary eosinophilia, acute eosinophilic pneumonia, chronic eosinophilic pneumonia, Churg Strauss syndrome, allergic bronchopulmonary aspergillosis, and idiopathic hypereosinophilic syndrome. Secondary eosinophilic lung diseases include all those caused by drugs, parasites, and fungi. The third group consists of systemic diseases associated with pulmonary eosinophilia such as sarcoidosis, malignant disease, and Langerhans cell histiocytosis. Our patient fulfi lled all criteria for acute eosinophilic pneumonia (table) and we had excluded other causes of pulmonary eosinophilia. A recent change in smoking habits is associated with acute eosinophilic pneumonia, as are other toxins. We started our patient on prednisolone, and after 48 h her symptoms of coughing and chest pain resolved and we could stop oxygen supplementation. 8 days later she refused treatment, including aripiprazole, and was transferred back to the psychiatric clinic. When she was discharged 13 days later she smoked 20 cigarettes in one day. That night her symptoms returned, accom panied by peripheral eosinophilia (12%), so we could confi dently attribute the acute eosinophilic pneumonia to cigarette re-exposure. A rebound event after stopping prednisolone cannot be excluded but has never been described, by contrast with relapse of acute eosinophilic pneumonia. She restarted prednisolone, but refused to keep taking the prednisolone 2 days later when she no longer needed oxygen therapy. At follow-up in December, 2014, she was healthy and not smoking. Acute eosinophilic pneumonia should be included in the diff erential diagnosis of pneumonia, especially in patients who have recently started smoking.