Intended for healthcare professionals


Difficult asthma or Churg-Strauss syndrome?

BMJ 1999; 318 doi: (Published 20 February 1999) Cite this as: BMJ 1999;318:475

Steroids may be masking undiagnosed cases of Churg-Strauss syndrome

  1. David P D'Cruz, Senior lecturer in rheumatology (D.P.Cruz{at},
  2. Neil C Barnes, Consultant physician in respiratory medicine,
  3. C Martin Lockwood, Reader in therapeutic immunology
  1. School of Clinical Medicine, Addenbrooke's Hospital, Cambridge CB2 2QQ

    The standard management of asthma with inhaled corticosteroids and bronchodilators works in most patients. A significant minority, however, respond poorly to these medications and require frequent intermittent courses of oral corticosteroids or occasionally long term oral corticosteroid treatment. The success of the new leukotriene antagonists in reducing the need for steroids in these patients with difficult asthma seems also to be unmasking some undiagnosed cases of Churg-Strauss syndrome, the symptoms of which were apparently being controlled by steroids.

    The pathophysiology of bronchoconstriction is not fully understood, but it is clear that leukotrienes, pro-inflammatory mediators arising from arachidonic acid, are important in the inflammatory processes of asthma. Several pharmacological agents, including the sulfidopeptide-leukotriene antagonists zafirlukast, montelukast, and pranlukast, have been studied in large scale randomised clinical trials.1 These trials have all been in patients with mild to moderate asthma who were not taking oral corticosteroids —or, if they were, the dose was not altered during the studies. Several cases of Churg-Strauss syndrome associated with the useof zafirlukast have been reported in the United States and more recently with montelukast in the United Kingdom.2 3

    Churg-Strauss syndrome is a rare allergic granulomatous eosinophilic vasculitis characterised by late onset asthma, upper airways disease including allergic rhinitis, sinusitis, and systemic vasculitis. Three phases to the disease have been described: firstly, an allergic rhinitis, often associated with recurrent sinusitis and asthma, that becomes progressively more difficult to treat; secondly, peripheral blood eosinophilia associated with eosinophilic pulmonary infiltrates and worsening asthma; and, thirdly, systemic vasculitis, commonly including peripheral neuropathies and occasionally life threatening cardiac disease.4 Diagnosis is based mainly on clinical grounds since eosinophilic granulomas are not always present on biopsy even in florid cases.4

    Though most patients with the syndrome present with fairly classic clinical features, atypical or formes frustes presentations are documented.5 Despite a good response to immunosuppressive treatment, many patients continue to suffer from debilitating asthma and upper airways disease, especially sinusitis. Most cases of Churg-Strauss syndrome are idiopathic, though inhaled allergens, vaccination, and desensitisation might be triggering factors.6 Drugs such as penicillin, sulphonamides, anticonvulsants, and thiazides have also been implicated. 3 7

    Wechsler et al reported eight patients with corticosteroid dependent asthma receiving zafirlukast who developed Churg-Strauss syndrome associated with withdrawal of oral corticosteroids.3 In retrospect, two of these patients had evidence of pre-existingChurg-Strauss syndrome, with asthma, peripheral neuropathy, or pulmonary infiltrates.3 In most of these patients zafirlukast improved asthma control and allowed tapering of the oral corticosteroids: Churg-Strauss syndrome developed, or became apparent, within days to months of stopping oral prednisolone. Prednisolone is an effective treatment for Churg-Strauss syndrome, and the prodromal phase of Churg-Strauss syndrome might have been partially treated in these patients by the steroids they were taking for their asthma.

    Patients with systemic vasculitis have an increased prevalence of allergies to drugs as well asto skin and airborne allergens,8 so a drug induced hypersensitivity vasculitis due to zafirlukast is possible, though drug induced Churg-Strauss syndrome is extremely rare. 37 A morelikely explanation is that zafirlukast improved the patients' asthma significantly enough to withdraw oral corticosteroids, unmasking an underlying vasculitis.3 The appearance of Churg-Strauss syndrome has certainly been documented in corticosteroid dependent asthma patients in whom the prednisolone dose was decreased or discontinued—two of these cases were fatal.5 Furthermore, the introduction of fluticasone, allowing tapering of the oral corticosteroid dose, led to the emergence of Churg-Strauss syndrome in one patient (N Barnes, unpublished observation). Another important factor is that in 58% ofpatients with Churg-Strauss syndrome the asthma improved spontaneously, possibly allowing corticosteroid withdrawal, just before the development of the vasculitic phase of the disease.4

    How is the general physician to distinguish early Churg-Strauss syndrome, where asthma is always a feature, from “difficult” idiopathic asthma? More importantly, can Churg-Strauss syndrome developing in an asthmatic patient who is stopping corticosteroids be recognised early enough to avoidsignificant morbidity and mortality? Clearly this is not easy, though there may be pointers. For example, upper airways disease, particularly sinusitis that is severe, recurrent, or requires surgery, may be a useful clue to Churg-Strauss syndrome, since this is not usually associated with simple allergic rhinitis. 4 9 10 Indeed, sinusitis with abnormal paranasal sinus radiographs has a high predictive value for Churg-Strauss syndrome.11

    Other useful pointers to a multisystem disease include joint pains, myalgias, malaise, neuropathy, palpable purpura, weight loss, or fevers.9 The development of such features before the introduction of oral corticosteroids, or on steroid tapering, shouldlead to further investigation. A persistent peripheral blood eosinophilia (especially if >1.5×109/l), transient pulmonary infiltrates or cardiomegaly on chest radiographs, microscopic haematuria, and raised erythrocyte sedimentation rates or C reactive protein values in the absence of infection should raise the suspicion of a vasculitis. Some patients with Churg-Strauss vasculitis have perinuclear staining antineutrophil cytoplasmic antibodies (p-ANCA) with specificity for myeloperoxidase, though a negative result does not exclude the diagnosis.6

    Thus in any patient with asthma (especially late onset asthma that is difficult to control) with features of a multisystem disease it is worth considering an underlying vasculitis that may be partially treated by oral corticosteroid therapy. When patients are already taking oral corticosteroids care should be taken in introducing other effective antiasthma agents, particularly leukotriene antagonists. These agents may improve asthma to the extent that oral corticosteroids can be withdrawn, unmasking a systemic vasculitis or possibly accelerating the disease to its life threatening vasculitic phase. As the oral corticosteroid dose is reduced such patients should be monitored carefully for the clinical appearance of a multisystem disease, with measurements of erythrocyte sedimentation rate, C reactive protein, and eosinophil counts.

    DD'C, NCB, and CML have provided independent expert opinions on vasculitis and asthma for Zeneca. NCB is involved in clinical trials of leukotriene antagonists for Zeneca, Ono, and Merck and has provided ad hoc consultancy advice and spoken at symposiums.