|Home | About | Journals | Submit | Contact Us | Français|
Angiotensin converting enzyme (ACE) inhibitors are responsible for several respiratory effects including cough and angioneurotic oedema but an association with pulmonary eosinophilic syndrome is uncommon.
A Caucasian woman aged 68 was seen after two weeks of malaise, nausea and sinusitis, for which she had been prescribed clarithromycin. Seven weeks earlier she had been started on perindopril for hypertension, at which time a full blood count was normal. She was also taking co-amilofruse and was using inhaled medications for asthma—fluticasone, begun eighteen months earlier to replace the oral prednisolone she had taken for 15 years, and salbutamol. There was no travel history of note and she kept no household pets. On examination she was apyrexial. No abnormal physical signs were detected—in particular no rash or subcutaneous nodules. Blood pressure was 109/80 mmHg and oxygen saturation 97% on air. Haemoglobin was 10.1 g/dL, MCV 93fL, neutrophils 9.96109/L and the eosinophil count was 13.46109/L (52%). Platelet count and clotting studies were normal; sodium was 128 mmol/L, potassium 4.8 mmol/L, urea 4.5 mmol/L, creatinine 58 μmol/L, albumin 31 g/L, corrected calcium 2.38 mmol/L, C-reactive protein 86 mg/L. There was no active urinary sediment. On sinus radiography the ethmoid and maxillary sinuses were opaque. A chest radiograph showed hyperinflated lungs and a small area of pleural scarring at the left costophrenic angle.
On admission to hospital, perindopril, clarithromycin and the diuretic were withdrawn, and she was continued on a fluticasone inhaler. A low-grade pyrexia developed and on day 6 the eosinophil count had risen to 18.36109/L (61%). CT revealed pulmonary infiltrates with mediastinal lymphadenopathy but no proximal bronchiectasis (Figure 1). Further investigations showed aspergillus precipitins weakly positive at titre 1:2, IgE 576 kU/L (normal 0-81), stool examination negative for ova, cysts and parasites, and echocardiogram normal. Serum angiotensin converting enzyme, extractable nuclear antibodies, antinuclear antibodies (ANA) and antineutrophil cytoplasmic antibodies (ANCA) were negative. On a short Synacthen (tetracosactide) test a baseline cortisol of 93 μg/dL rose to 119 μg/dL, indicating adrenal insufficiency. A bone marrow aspirate and trephine biopsy showed normal bone marrow precursors with no evidence of underlying malignancy.
After a bronchoscopy she became nauseated and hypotensive, and the eosinophil count rose further to 25.56109/L (72%). She was treated with a single dose of 100 mg intravenous hydrocortisone and within forty-eight hours the eosinophil count had fallen to 0.16109/L (51%). She was then started on oral hydrocortisone replacement therapy (20, 10, 10 mg per day).
A transbronchial biopsy showed a pronounced eosinophilic infiltrate but no evidence of vasculitis (Figure 2). A long Synacthen test suggested preserved adrenal function, but an undetectable baseline morning cortisol was in keeping with insufficiency at the level of the hypothalamopituitary axis. This was presumed secondary to previous long-term steroid use and she continues on oral hydrocortisone replacement therapy at doses of 10, 5 and 5 mg per day. Repeat CT of the thorax six months later was normal and autoantibody screening, adrenal antibodies, ANA and ANCA were persistently negative. At twenty-four months her asthmatic symptoms remained well controlled on fluticasone and the eosinophil count was normal.
Peripheral eosinophilia can be caused by allergic disorders, parasitic infections, lymphoproliferative disorders, solid tumours, gastrointestinal diseases, skin conditions, collagen vascular diseases and systemic vasculitides.1 Drug-induced syndromes should also be considered but only after exclusion of other diagnoses. ACE inhibitors are associated with cough, and hypersensitivity lung disease is increasingly recognized in connection with these agents. Eosinophilic syndromes have been attributed to captopril,2 enalapril3 and fosinopril.3 In these instances the eosinophilia was not as great as that reported here. The temporal relation of perindopril treatment and clinical features of hypersensitivity point to the causal involvement of this ACE inhibitor in our patient with eosinophilic pneumonia—an adverse reaction that has not previously been reported with this agent. Rechallenge with perindopril was considered too hazardous to be justifiable.
Other medications implicated in drug-induced pulmonary eosinophilic syndromes include antibiotics and non-steroidal anti-inflammatory agents.4 Terzano et al.5 report a patient with asthma who developed fever and pulmonary infiltrates with eosinophilia after taking clarithromycin, on two separate occasions, responding to withdrawal of the antibiotic and institution of corticosteroid treatment. Our patient received clarithromycin before presentation but has since been unintentionally rechallenged without recurrence of pulmonary infiltrates or eosinophilia.
It is possible, in view of her history of asthma, that she had an underlying vasculitic process that was partly controlled by the inhaled fluticasone; systemic absorption of inhaled steroids is recognized.6 We feel that the absorbed pulmonary steroid combined with the current oral steroid dose would be insufficient adequately to control a vasculitis. Her ANCA was also negative and there were no other clinical features to support a diagnosis of vasculitis. The diagnosis of Churg-Strauss syndrome should be suspected in any patient presenting with eosinophilia, asthma, pulmonary infiltrates and evidence of vasculitis. The asthma is usually of late onset and may become more severe with time.7 Extrapulmonary features can be helpful in distinguishing Churg-Strauss from chronic eosinophilic pneumonia and allergic bronchopulmonary aspergillosis: in particular Churg-Strauss manifests, in up to 75% of patients, with mononeuritis multiplex.7 At twenty-four months our patient remained free from neurological symptoms, and had no evidence of cardiac, skin, joint or renal involvement. Hypereosinophilic syndrome may represent a preleukaemic state, or may herald the onset of Hodgkin's disease or T-cell lymphoblastic lymphoma.8 An underlying lymphoproliferative or myeloproliferative disorder is unlikely in the present case in view of the normal bone marrow, rapid response to hydrocortisone and lack of recurrence of the peripheral eosinophilia. Mild eosinophilia associated with adrenocortical insufficiency is well recognized in critically ill patients. However, the pulmonary infiltrates, mediastinal lymphadenopathy and marked eosinophilia seen in our patient cannot be explained by hypocorticism. Furthermore, the strong IgE response suggests a hypersensitivity phenomenon such as a drug reaction.
We thank Dr Jeremy Berger for his assistance with diagnostic radiology, Dr Andrew Gallimore for providing histopathology images and Dr Adrian Barnes for his helpful comments.