Diagnosis of Churg-Strauss syndrome: eotaxin-3 makes it easier

This editorial refers to ‘Eotaxin-3 in Churg-Strauss syndrome: a clinical and immunogenetic study’, by Jochen Zwerina et al., doi:10.1093/rheumatology/keq445, on pages 1823–1827. The Churg–Strauss syndrome (CSS) is defined as a systemic necrotizing vasculitis affecting small-to-medium-sized vessels; CSS is also characterized by tissue eosinophilia, eosinophilic extravascular granulomas, peripheral eosinophilia, asthma, allergic rhinitis and sinusitis. In 1951, Churg and Strauss were the first to correlate these pathological and clinical findings, and defined this syndrome as allergic granulomatosis and angiitis [1]. Churg’s and Strauss’s definition of granulomatosis was based on the pathological observation of allergic granulomas in different organs. The allergic or eosinophilic granuloma, however, is not pathognomonic of CSS, since it can also be found in other systemic vasculitides, connective tissue disorders, lymphoproliferative diseases, subacute bacterial endocarditis, chronic hepatitis and ulcerative colitis. Equally, tissue infiltration by eosinophils is also observed in a variety of other conditions including the hypereosinophilic syndrome, drug hypersensitivity, parasitic infections, malignant tumours and various allergic diseases. The ACR criteria [2], which include clinical and laboratory findings such as asthma, eosinophilia >10%, peripheral neuropathy, migratory lung infiltrates and sinusitis, can only be used for the classification of a vasculitis syndrome as CSS, with a sensitivity of 85% and a specificity of 99.7% when at least four of them are present; these criteria, however, cannot be used for the diagnosis of CSS. The Chapel Hill Consensus Conference (CHCC) definitions [3] were mainly descriptive, but are neither suitable for the diagnosis nor for the classification of CSS. Nevertheless, given the lack of validated diagnostic tools, the ACR criteria and the CHCC definitions are often mistakenly given diagnostic significance. The classical evolution of CSS into three distinct phases provides, in some cases, a tool for the diagnosis. The prodromic phase, which can persist for many years, is hallmarked by allergic asthma (usually with an adult onset) in patients who are often not historically atopic, as well as by allergic rhinitis and nasal polyps, which tend to recur after surgery. The second phase is characterized by peripheral eosinophilia and tissue eosinophilic infiltrates that can affect the respiratory system, the gastrointestinal tract and the myocardium. In the third phase, constitutional symptoms and clinical manifestations due to vasculitic organ involvement predominate. During this phase, allergic manifestations may persist (albeit they are often attenuated), but the clinical picture essentially features peripheral neuropathy, skin lesions and GN. The sequence of phases is not always so clear cut, they may overlap or some of them may not be present at all, so they can be helpful to the clinician only when they are found in the typical order. Among the laboratory tests that support the diagnosis of CSS, blood eosinophilia (>10% or >1500 cells/µl) is extremely frequent. Increased CRP levels and ESR are common as well as anaemia. Total serum immunoglobulin E (IgE) levels are high in most cases, but IgE specific to common allergens is positive in only a few patients [4]. We can also find increased levels of eosinophilic cationic protein and, less frequently, RF [5]. In this issue of Rheumatology, Zwerina et al. [6] bring an effective contribution to the diagnosis of CSS in its active phase. Thirty-seven patients with a diagnosis of CSS (active, relapsed or inactive), 123 healthy controls and 138 disease controls were studied and serum levels of eotaxin-3 were determined. The authors show that eotaxin-3 was significantly higher in active CSS, and that it clearly differentiated active CSS from inactive CSS and, more importantly, from all the disease control groups. The strength of this work lies in the inclusion of a wide range of control groups (i.e. other vasculitides, CTDs, hypereosinophilic syndromes, parasitic diseases, ulcerative colitis), which strengthens the significance of eotaxin-3 in the differential diagnosis of CSS: in this setting, eotaxin-3 showed very high specificity and sensitivity (98.6 and 85.5%, respectively), which is why it can become a suitable marker for routine clinical practice. CSS is classified among the so-called ANCA-associated systemic vasculitides, although ANCA positivity (where ANCA is usually directed against MPO) occurs in only 40% of CSS patients. ANCA is also a valuable tool in differentiating CSS from other non-vasculitic diseases, but its sensitivity is low. ANCA may help to distinguish two subsets of CSS. ANCA-positive CSS often features mononeuritis multiplex, alveolar haemorrhage, necrotizing crescentic GN and purpura, all of which are ascribed to small-vessel vasculitis. ANCA-negative CSS is instead associated with myocarditis, pleural effusion, pulmonary infiltrates and nasal polyposis, which are mainly due to eosinophilic infiltration and eosinophil toxic enzyme release into the involved tissues. Some clinical manifestations such as constitutional symptoms, hypereosinophilia, asthma, gastritis and/or enteritis, mono-polyneuropathy and CNS involvement have similar distribution in the two subsets [7, 8]. In the paper by Zwerina et al. [6], eotaxin-3 levels did not differentiate between these subsets; this might indicate that increased eotaxin-3 production relates to disease mechanisms that are common to both ANCA-positive and ANCA-negative CSS subsets. Some genetic features can potentially be helpful for the diagnosis of CSS: HLA-DRB4 [9] is a well-demonstrated genetic risk factor for the development of CSS and is probably linked to the vasculitic manifestations of the disease; conversely, the IL-10.2 haplotype is significantly associated with ANCA-negative CSS [10]. However, these genetic polymorphisms have not been systematically evaluated in other disease groups; therefore, their role in the differential diagnosis is still uncertain. The main clinical, laboratory and histological features of CSS and their significance in the differential diagnosis with other diseases are summarized in Table 1. Differential diagnosis of CSS on the basis of its main clinical, laboratory and histological findings The frequency of the listed features is graded from +++ (very frequent) to − (usually absent). HES: hypereosinophilic syndrome; AASV: ANCA-associated systemic vasculitides [which includes WG, and microscopic polyangiitis (MPA)]; CTD (which includes SLE and SSc); BAL: broncho-alveolar lavage; RPGN: rapidly progressive GN. Differential diagnosis of CSS on the basis of its main clinical, laboratory and histological findings The frequency of the listed features is graded from +++ (very frequent) to − (usually absent). HES: hypereosinophilic syndrome; AASV: ANCA-associated systemic vasculitides [which includes WG, and microscopic polyangiitis (MPA)]; CTD (which includes SLE and SSc); BAL: broncho-alveolar lavage; RPGN: rapidly progressive GN. In conclusion, the diagnosis of CSS may be challenging even if histological proof is available; medical history, clinical features and laboratory findings need to be integrated to reach a correct diagnosis; the availability of new disease-specific markers such as serum eotaxin-3 represents a great addition to the diagnostic armamentarium and warrants validation in clinical practice. Early diagnosis of CSS may definitely allow early treatment, thus preventing the development of irreversibile organ damage. Disclosure statement: The authors have declared no conflicts of interest.