Answer

From the question on page 214

Churg-Strauss syndrome (CSS) was diagnosed on the basis of extensive vasculitis in a sural nerve biopsy, tissue eosinophilia in a cecum biopsy (see figure in online supplement) and peripheral eosinophilia (11%, observed only once).1 The cecum biopsy was performed because anaemia had become microcytic; pathological analysis was consistent with an eosinophilic gastroenteritis. p-Antineutrophil cytoplasmic (titre 1:1280) and anti-myeloperoxidase (657 U/ml) antibodies were strongly positive. CSS disease activity markers soluble interleukin-2 receptor (7748 pg/ml) and eosinophilic cationic protein (13.2 mg/l) were elevated.2 There was no history of the use of antileukotriene agents or acetaminophen. Although no single CT image is pathognomonic for CSS, ground-glass attenuation appears to be the most consistent feature.3 Pulmonary fibrosis is not usually mentioned as a complication of CSS1 but has been reported previously4 and may represent the final common pathway of untreated disease.

The following points can be learnt from this case. CSS can present with symptoms resembling rheumatoid arthritis5 but joint involvement is not usually mentioned as a manifestation of CSS.1 CSS may be present without a formal diagnosis of asthma because pulmonary vascular disease can also be characterised by low diffusing capacity without obstruction.6 Limited testing for polyneuropathy in the context of diabetes mellitus is usually recommended.7 However, severe pain, asymmetry and predominant or progressive motor symptoms (eventually all present in this patient) should serve as red flags and warrant further testing.7 Early recognition of CSS is important because treatment may stop the potentially dangerous progression of disease and reverse symptoms. In that regard, the presence of allergic rhinitis should have served as a diagnostic clue. Continuous low-dose steroid treatment may have disguised other diagnostic clues such as eosinophilia in blood and lavage fluid. Instead, a bronchoscopic or surgical lung biopsy early in the course may have expedited diagnosis. This case vividly demonstrates one of medicine's challenges—the necessity to search for all possible diagnoses to avoid missing one, but also to avoid missing a single unifying diagnosis during this search. This case also illustrates the pitfalls of subspecialisation. The rheumatologist, respiratory physician and neurologist found alternative explanations for symptoms of CSS within their areas of expertise, obscuring the larger picture.

The patient was treated with prednisone (30 mg daily) and azathioprine (100 mg daily). Fever did not return, polyneuropathy improved and erythrocyte sedimentation rate, C-reactive protein and haemoglobin normalised.