Conflict of interest: none declared.
Clinical dermatology •Concise report
Persistent hypereosinophilia with Wells syndrome
Article first published online: 21 MAY 2012
© The Author(s). CED © 2012 British Association of Dermatologists
Clinical and Experimental Dermatology
Volume 38, Issue 1, pages 40–43, January 2013
How to Cite
Powell, J., Kaur, M., Muc, R., Colloby, P. and Salim, A. (2013), Persistent hypereosinophilia with Wells syndrome. Clinical and Experimental Dermatology, 38: 40–43. doi: 10.1111/j.1365-2230.2012.04370.x
- Issue published online: 18 DEC 2012
- Article first published online: 21 MAY 2012
- Accepted for publication 16 January 2012
Vol. 38, Issue 3, 330, Article first published online: 21 MAR 2013
Since Wells and Smith first described cases of eosinophilic cellulitis (Wells syndrome; WS) in 1979, it has been noted that some but not all patients with WS present with eosinophilia. In the face of idiopathic persistent eosinophilia patients will also then fall within the hypereosinophilic syndrome (HES), which represents a multifarious spectrum of disorders of varying severity, causes and outcomes. In this article we propose that patients who present within the HES spectrum with cutaneous findings of WS and with no extracutaneous disease be classified as having ‘persistent hypereosinophilia with Wells syndrome’ (PHEWS).
In 1979, Wells and Smith described cases of eosinophilic cellulitis (also known as Wells syndrome; WS). Since then, it has been noted that some patients with WS present with eosinophilia. In the face of idiopathic persistent eosinophilia, patients may also then fall within the description of hypereosinophilic syndrome (HES) which represents a multifarious spectrum of disorders of varying severity, causes and outcomes. We present the case of a patient with WS who also falls within the HES but with no extracutaneous manifestations and propose a new term and diagnostic criteria for this condition.
A 64-year-old woman presented with a 6-year history of a pruritic skin eruption. Individual lesions persisted for several weeks before fading, leaving postinflammatory hyperpigmentation, but crops of new lesions continued to arise. The patient remained systemically well throughout.
On physical examination, red papules, nodules and urticarial-like lesions were present on the patient’s trunk and limbs (Fig. 1a).
On histological examination, the epidermis was normal in appearance. Perivascular and interstitial infiltrates were seen within the superficial dermis, containing numerous eosinophils and occasional flame figures (Fig. 1b).
Laboratory investigations identified hypereosinophilia, which had been present throughout the illness, peaking at 4.12 × 109 eosinophils per litre (normal range 0–0.4 × 109/L), and being persistently above 1.5 × 109/L for 5 years. Full blood count, renal and liver function tests, autoimmune profile, and measurement of erythrocyte sedimentation rate, C-reactive protein, serum IgE and tryptase levels gave normal results. Tests for filarial, schistosomal and strongyloides serology were negative. No abnormalities were seen on chest radiographs. Cytogenetic and molecular studies did not identify any clonal haematological abnormality, including FIP1L1/PDGFRA (family-interacting protein 1-like 1/platelet-derived growth factor receptor-α) gene fusion.
Three years previously, a short course of oral steroids had brought about a near-complete clinical remission and had markedly improved the eosinophil count, but a full relapse occurred shortly after the steroids were stopped.
In this case, a clinicopathological diagnosis of WS could be made, however, we noted that the patient also falls within the HES spectrum.
In their paper, Wells and Smith1 described the characteristic clinocopathological features of WS. Classically and in its most severe form, WS presents initially with an acute pseudocellulitis, which may extend to affect the greater part of a limb. Areas are red and oedematous, and may blister, but the affected skin does not have a raised temperature on palpation, and is unresponsive to antibiotics. Lesions then become infiltrated, firm and granulomatous, with green-blue discoloration, resolving over several weeks, during which lesions may resemble morphoea.1
Wells and Smith then described other, less severe, cutaneous presentations of WS, namely those with urticaria-like, bullous, papular, nodular, annular and plaque-like lesions, which are often pruritic. In the cases they presented, peripheral blood eosinophilia was not always present, and no comment was made on how long any eosinophilia (if present) persisted, or if this recurred in the future.
Such multifarious clinical findings continue to be reported in patients diagnosed with WS, meaning that the diagnosis will sometimes be made solely on characteristic cutaneous histological features.1,2 Such features include dermal oedema with infiltration of eosinophils and macrophages between connective-tissue bundles, with scattered ‘flame figures’ (clusters of eosinophils and macrophages around a central focus of collagen coated in granular eosinophilic material).
Wells and Smith stated originally that WS is probably a hypersensitivity reaction induced by various triggers such as insect bites or drug reactions.1 Since their study, WS has been reported to occur in systemic disorders associated with eosinophilia such as Churg–Strauss syndrome,3 HES4 and chronic lymphocytic leukaemia.5 It therefore seems that WS represents a cutaneous reaction pattern that can develop in any situation in which abnormal numbers of eosinophils are allowed to accumulate in the skin for any, often unknown, reason.
Hypereosinophilic syndrome is a term covering a wide range of heterogenous conditions. Since 1975, three criteria have been used to define HES: (i) blood eosinophilia over 1.5 × 109/L for > 6 months or death within that time, associated with signs and symptoms of hypereosinophilic disease; (ii) no secondary cause for eosinophilia found; and (iii) signs and symptoms of organ damage.6 In HES, tissue eosinophilic infiltration and mediator release leads to variable end-organ damage, which despite aggressive treatment, may result in severe complications and even death.7 Such complications include cardiac failure and fibrosis, peripheral neuropathy, and generalized central nervous system dysfunction such as upper motor neurone signs and encephalopathy.7 Pulmonary features (such as fibrosis and nonproductive cough), hypersplenism, hepatitis, hepatic-vein obstruction, eosinophilic oesophagitis, gastritis or colitis, and choroidal abnormalities can all occur.7 However, some patients with HES are asymptomatic or present with more ‘benign’ disease, perhaps limited to the skin, and these patients may never, despite chronic hypereosinophilia, develop such complications.
Since it was first described, the underlying aetiologies in some cases of HES have been identified. For diagnostic purposes, there is no longer the insistence that the hypereosinophilia be ‘truly’ idiopathic, so long as secondary causes of eosinophilia, such as drug-induced, parasitic, allergic and autoimmune causes, have been excluded.7,8 For example, patients with myeloproliferative (M)-HES can now be identified through molecular and genetic techniques. The commonest chromosomal mutation seen in M-HES is the deletion on chromosome 4q12, causing the fusion of the FIP1L1/PDGFRA (F/P) genes, leading to a protein with significant tyrosine-kinase activity. Furthermore, using flow cytometry, lymphocytic (L)-HES has been identified, characterized by an aberrant population of T-cells of uncertain origin, which secrete eosinophilopoietic cytokines [interleukin (IL)-3 and/or IL-5] causing eosinophilia.8,9
Owing to this ongoing aetiological uncertainty in many cases of HES, an attempt has been made to classify patients within the spectrum based on their clinical and pathological features. In 2005, six subtypes of HES were proposed:8 (i) M-HES (including F/P negative with myeloproliferative features, F/P positive and chronic eosinophilic leukaemia with cytogenetic abnormalities); (ii) L-HES (with demonstrable clonal or phenotypically aberrant lymphocyte population); (iii) familial eosinophilia (including an autosomal dominant form mapped to chromosome 5q31–33); (iv) unidentified HES (idiopathic with or without symptoms, including episodic variants); (v) overlap HES (eosinophilic disease restricted to one organ system with eosinophilia > 1.5 × 109/L); and (vi) associated HES (eosinophilia > 1.5 × 109/L in the setting of another diagnosis, in which eosinophilia has been described as a feature in a subset of those affected).8,9
Identification of different subtypes of HES is important to allow treatment to be tailored and long-term outcomes predicted in individual cases within the HES spectrum. For example, corticosteroids were initially the first-line treatment for all types of HES, but since the identification of the F/P mutation, these patients are now treated successfully with imatinib mesylate.
Cutaneous involvement is present in 45–60% of patients with HES, and pruritus, erythematous papules, nodules, urticara-like lesions and angio-oedema have all been described.7,10,11 Diagnoses of WS occurring in such patients with HES continue to occur, and include the case we describe above.4 Wells and Smith1 recognized this overlap in 1979, and regarded those affected as being at the ‘benign end’ of the HES spectrum (‘benign’ here indicating no extracutaneous complications resulting from eosinophilia). However, time has shown that the term ‘benign’ is misleading, as some patients with chronic eosinophilia can and will develop internal organ disease as detailed above, including transformation to leukaemia or lymphoma.7
Using current classification schemes, those with WS and persistent idiopathic eosinophilia but no extracutaneous organ involvement do not fit happily into any single subtype of HES. Furthermore, eosinophilia, although often present, is not needed to make a diagnosis of WS, and therefore patients should not be regarded as having ‘solely’ WS, especially if the eosinophilia is persistent. It is important to be able to identify and classify such patients on clinicopathological grounds, so that the aetiology, treatments and outcomes in such cases can be defined.
We have therefore set out the case for distinguishing patients with WS from those with WS plus persistent idiopathic eosinophilia with no extracutaneous organ involvement. Such cases are a readily identifiable subgroup of patients within the HES spectrum. Currently no classification system adequately identifies such patients, and thus we propose a new classification of persistent hypereosinophilia with Wells syndrome: ‘PHEWS’, using the following diagnostic criteria: (i) idiopathic persistent hypereosinophilia > 1.5 × 109/L; (ii) clinicopathological features of WS; and (iii) no eosinophilia-mediated extracutaneous organ involvement.
‘Idiopathic’ here means that any causes of M-HES and secondary eosinophilia have been excluded. This will include referral for a haematological assessment, including measurement of serum tryptase levels, FIP1L1/PDGFRA analysis, and possibly bone-marrow biopsy with cytogenetic, molecular, histochemical and immunophenotypic studies.11 We have used the cut-off point of > 1.5 × 109/L for > 6 months, as used in the HES diagnostic criteria established by Chusid et al.6 in 1975, to define ‘persistent hypereosinophilia’. When carrying out investigations for any extracutaneous complications of persistent hypereosinophilia, physicians can be guided by clinical findings and symptoms, with a low threshold of suspicion for cardiac and pulmonary involvement.
Treatment in cases of confirmed PHEWS will largely be aimed at symptomatic relief, and for now remains as for WS, with systemic corticosteroids used initially, and dose adjustments according to response. Topical corticosteroids alone may be sufficient. Antihistamines, dapsone, ciclosporin, azathioprine, griseofulvin, tetracyclines and interferon-α2a have also been reported to be effective and useful as steroid-sparing agents in patients needing long-term treatment.12 Currently, as with many issues around HES, patient outcome, treatment goals and disease progression are largely uncertain in PHEWS. Monitoring such patients with persistent hypereosinophilia on a long-term basis to some degree seems currently sensible, as it is not yet possible to predict what, if any, damage such persistent hypereosinophilia may cause or whether a haematological malignancy will develop in the future. We hope that with the establishment of PHEWS as a distinct syndrome, such questions will be answered in the future.
- •Wells syndrome is a reactive process in which some patients present with eosinophilia and may therefore be within the HES spectrum.
- •Any eosinophilia present in the context of WS must be thoroughly investigated, especially if persistent.
- •Patients can be diagnosed with ‘PHEWS’ if they have idiopathic, persistent hypereosinophilia (>1.5 × 109/L) plus the clinicopathological features of WS with no evidence of any eosinophil-mediated extracutaneous organ involvement.
- •Identifying such patients is important so long-term outcomes, treatment goals and management strategies can be defined.