Churg-Strauss syndrome (CSS) is a systemic necrotizing vasculitis affecting small to medium-sized vessels, and characterized by asthma, blood hypereosinophilia, and eosinophil-rich granulomatous inflammation of the respiratory tract. In the past few years the pathogenesis of the disease and its clinical manifestations have been clarified, fostering important advances in the treatment of CSS. Systemic corticosteroids are still considered the cornerstone of treatment. Many issues need to be addressed, such as how to maintain remission, prevent disease relapses, and treat refractory disease. This review provides a clinical overview of CSS and a summary of the current treatments and novel therapies.
Churg-Strauss syndrome (CSS) is a systemic necrotizing vasculitis named after the two pathologists, Churg and Strauss, who first described the disease in 1951. In their original work, Churg and Strauss described 13 patients presenting with a clinical syndrome characterized by asthma, hypereosinophilia, and evidence of vasculitis affecting several extrapulmonary organs and identified the disorder as a separate disease distinguishable from panarteritis nodosa and from the other systemic vasculitides. Over time, the clinical picture of CSS has been progressively delineated and classification/diagnostic criteria have been elaborated to better define the particular aspects of the disease. In 1990, the American College of Rheumatology developed a set of diagnostic criteria for the disease, including asthma, eosinophilia greater than 10% on differential white blood cell count, mononeuropathy (including multiplex mononeuritis multiplex) or polyneuropathy, nonfixed pulmonary infiltrates on roentgenography, paranasal sinus abnormality, and biopsy containing a blood vessel with extravascular eosinophils. The presence of 4 or more of these 6 criteria yielded a sensitivity of 85% and a specificity of 99.7%. More recently, the Chapel Hill Consensus Conference defined CSS as an eosinophil-rich and granulomatous inflammation involving the respiratory tract and a necrotizing vasculitis affecting small to medium-sized vessels, associated with asthma and eosinophilia. Finally, CSS has been classified along with Wegener’s granulomatosis (WG) and microscopic polyangiitis (MPA) as one of the antineutrophil cytoplasmic antibodies (ANCAs)–associated vasculitides because of the presence of ANCAs, which have been documented in approximately 40% of patients. So far, a few large clinical series have been published assessing the clinical features of the disease and the long-term outcome of CSS patients. In all these series CSS has been described as a complex systemic disorder sharing the features of a systemic small-vessel vasculitis and those of an allergic hypereosinophilic disorder, with vasculitis and hypereosinophilia variously involved in the pathogenesis of the different manifestations of the disease. From a therapeutic point of view, there is a wide consensus that the use of systemic corticosteroids, in combination with immunosuppressants for the most severe cases, seems to be the effective cornerstone of the treatment of the disease, which has a good prognosis compared with other small vessel systemic vasculitides. This review provides a clinical overview of CSS and a summary of the current treatment as well as novel therapies for the disease.
What have we learned from large clinical series?
CSS is a rare disorder characterized by an incidence of 0.11 to 2.66 new cases per million population per year and an overall prevalence of 10.7 to 14 per 1,000,000 adults. The male-to-female ratio ranges from 0.3 to 2.3 and the mean age at the onset reported in the literature varies from 38 to 52 years (range 7–74 years). Due to its rarity, the number of published large clinical series of CSS is limited. Nonetheless, there is a general agreement in the clinical description of the disease among the different series. The reported frequencies of organ manifestations and clinical features of CSS are summarized in Fig. 1 .
CSS is usually described as a 3-stage disorder characterized by a prodromal phase dominated by asthma, rhinosinusitis, and nasal polyposis followed by a second phase characterized by peripheral blood and tissue eosinophilia, and, finally, a third, properly defined, vasculitic phase. Although these 3 phases do not necessarily have to follow one another, all the largest clinical series justified their patients’ clinical and serologic features on the pathogenetic background of these stages of the disease. Among the possible clinical manifestations of the disease, a few features have been recognized as distinctive for CSS and are considered able to differentiate CSS from the other systemic vasculitides and from the other hypereosinophilic syndromes (HESs).
Late-Onset Asthma and Other CSS Distinguishing Features
Asthma seems to be the unifying feature of all CSS patients (96%–100%) and it is the main feature of the prodromal phase. Anecdotally, the development of asthmatic symptoms after the onset of the vasculitic phase has also been reported, but it represents the exception and not the rule. The typical latency between asthma and the vasculitic phase of the disease is estimated to last for an average of 3 to 9 years but has been reported as long as 30 years. This latency could be explained in part by treatments used for asthma control. For example, oral corticosteroids might partially suppress full-blown vasculitis for long periods of time, leading to forme frustes of CSS, which are unrecognized until corticosteroid tapering. The relationship between leukotriene receptor antagonists and other antiasthmatic drugs with the development of CSS is under debate. Although some investigators believe these drugs play a causative role in CSS, some believe that antileukotrienes and other drugs, such as the anti-IgE monoclonal antibody, omalizumab, simply make it possible for the patients to taper corticosteroids, thereby unmasking the underlying vasculitic process.
In comparison with patients with common atopic asthma, CSS patients typically have an adult-onset asthma, which usually becomes more severe with time and is often refractory to the traditional inhalation treatment. Up to three-quarters of patients require oral corticosteroids for adequate asthma control before the diagnosis of CSS. Asthma is often associated with moderate or severe blood eosinophilia. Airway inflammation in CSS may be severe, because eosinophil levels in induced sputum are higher than in asthma.
Asthmatic symptoms might exacerbate, but more commonly decrease, when the vasculitic phase arises. Nonetheless, asthma tends to be a persistent manifestation of the disease and is one of the most important factors that negatively influence the quality of life of the patients, even when the other disease manifestations are under therapeutic control.
Asthma is frequently associated with allergic rhinitis (47%–93%), nasal polyps, obstruction, and chronic or recurrent paranasal sinusitis (62%–77%) as seen on radiographs or CT scan. Nasal polyps, in particular, are widely considered a distinct feature of the disease and may cause the chronic airway obstruction, which is a classic manifestation of damage in CSS. Unlike WG, however, the presence of nasal pain, crusting, or hemorrhage is uncommon, and, overall, granulomas with eosinophilic infiltrates are less common than symptoms attributable to allergic rhinitis.
The role of common allergens in the prodromal phase of CSS as a common background for asthma and allergic rhinitis is debated. Recently, evidence of respiratory allergy, as demonstrated by specific IgE consistent with the clinical history, was found in less than one-third of 51 consecutive unselected CSS patients. This supports the hypothesis that allergy might be only one of the several mechanisms triggering exacerbation of asthma or upper airways symptoms in CSS.
In addition to upper airway involvement, the lung is a typical target organ in CSS. Pulmonary symptoms occur in 37% to 77% of patients and represent the result of a vasculitic process combined with varying degrees of eosinophilic infiltration. Transient pulmonary infiltrates with asthma and eosinophilia precede systemic vasculitis in 40% of cases but can occur in the prodromal phase or the vasculitic phase. During the first 2 phases, the histologic picture is generally that of extensive eosinophilic infiltration, whereas during the latter phase, necrotizing vasculitis and granulomas are more common. Bronchial alveolar lavage usually reveals a normal total cell count but with a dramatic increase in the percentage of eosinophils. Radiologically, infiltrates are generally patchy, asymmetric, or diffuse, without lobar or segmental distribution. Bilateral nodular infiltrates can also occur, but in contrast to WG, they rarely cavitate. The most common abnormalities on high-resolution CT scanning consist of bilateral ground-glass opacities and bronchial wall thickening. Infiltrates can be associated with pleural effusion in approximately 20% to 30% of cases.
Another distinctive feature of CSS is the involvement of the neurologic system. Peripheral neuropathy is the most important feature of the vasculitic process in CSS, occurring in up to 75% to 81% of patients. Neurologic symptoms may also be the presenting symptoms in approximately 20% of CSS patients. The vasculitic process typically involves the peripheral nerves, resulting in mononeuritis multiplex. Patients may present with sudden weakness or foot or wrist drop along with sensory deficits in the distribution of one or more distal nerves. In some patients, distal symmetric sensorimotor peripheral neuropathies have also been described. The common peroneal nerve and the internal popliteal nerve are the most frequently involved nerves; nonetheless, nerves of the upper limbs (ie, ulnar and radial nerves) might be involved as well. Sural nerve biopsy is the gold standard test in documenting the involvement of peripheral nervous system (PNS) in systemic vasculitis, but pathologic confirmation of necrotizing vasculitis and perineural infiltration of eosinophils and inflammatory cells have been reported only in half of the cases, with axonal degeneration the most common finding. The severity of neurologic manifestations varies, but even when severe, neurologic symptoms usually tend to respond promptly to the standard therapy. Some years ago, the authors described the case of a patient with an extremely severe peripheral involvement, tetraparesis, complete loss of strength (the patient was unable to move even in the absence of gravity), and a complete denervation pattern on electroneurography. The patient recovered almost completely after immunosuppressive therapy with only a mild residual motor deficit. Similar data are reported by others, emphasizing that PNS involvement may be dramatic but usually tends to resolve. Among residual minor symptoms, hyperesthesia or neuropathic pain, sometimes associated with mild motor deficit, seems common. Central nervous system involvement is less frequent but has been associated with a significant morbidity and mortality. Central vasculitis may predispose to hemorrhagic cerebrovascular events, especially in association with uncontrolled hypertension. Cranial nerve palsies are infrequent, but involvement of cranial nerves II, III, VII, and VIII has been described. The most common cranial nerve lesion is an ischemic optic neuritis.
Other Clinical Manifestations of the Eosinophilic and Vasculitic Phases of CSS
The early stages of CSS are often marked by general nonspecific constitutional symptoms consisting of malaise, significant and rapid weight loss (>5% body weight), and fever. Diffuse myalgia and polyarthralgia have been also reported in 37% to 57% of CSS patients, especially at the onset of the disease. Polyarthralgias have been described as mostly migratory and involving all the joints. Arthritis and synovitis have been reported in a smaller percentage of patients (<20%).
More than half of patients (53%–68%) experience cutaneous manifestations, which reflects the predominant vasculitic involvement of small vessels. Palpable purpura on the lower extremities is the most frequent skin lesion along with subcutaneous skin nodules of the scalp and of the limbs, urticarial rashes, cutaneous infarction, and livedo reticularis. The skin is the most commonly biopsied tissue due to the ease of biopsy. Although there are no findings specific for CSS, skin biopsy frequently allows clinicians to confirm the diagnosis in the appropriate clinical setting.
In their original cohort of patients, Churg and Strauss found myocardial abnormalities in more than 50% of the autopsies, ranging from the extensive replacement of myocardium by granulomas and scar tissue to coronary vessel vasculitis. Myocardial damage might be caused by toxic mediators released by activated infiltrating eosinophils or by vasculitic lesions in myocardium and in the coronary vessels. Myocarditis may lead to postinflammatory fibrosis and restrictive cardiomyopathy or congestive cardiac failure whereas coronary vasculitis may result in ischemic heart disease. The spectrum of cardiac manifestations includes acute pericarditis, constrictive pericarditis, restrictive or dilated cardiomyopathy, myocarditis, arrhythmias, and sudden death. When these manifestations are present, they are associated with a worse prognosis, accounting for approximately 50% of deaths. Considering the potential adverse outcomes associated with heart involvement in CSS, early detection is of clinical importance. Recently, cardiac MRI (CMRI) has been proposed as a useful tool to characterize myocardial involvement in CSS patients with or without clinical symptoms,and is purported to identify histologically proved fibrosis and active myocarditis. The frequency of abnormalities on CMRI is concordant with the reported percentage of myocardial involvement found during autopsy examinations of CSS patients and suggests a high incidence of cardiac manifestations in CSS patients even during remission. It is, however, debated whether or not CMRI should be performed in asymptomatic patients given the high frequency of false-positive CMRI results.
Digestive tract symptoms, including abdominal pain, intestinal obstruction, nausea, vomiting, diarrhea, and bleeding are common in CSS patients, reported in up to one–third of cases. From a pathogenetic point of view, these gastrointestinal manifestations may be related to 2 different mechanisms: eosinophilic infiltration of the bowel wall and mesenteric vasculitis. Eosinophilic intestinal infiltration may mimic most of the other forms of eosinophilic gastroenteritis and may precede or coincide with the vasculitic phase of CSS. Submucosal infiltrates can produce obstructive nodular masses, whereas mucosal involvement may result in diarrhea and bleeding. Serosal disease can produce an eosinophilic peritonitis with ascites with the ascitic fluid typically containing large numbers of eosinophils. Small and medium-sized mesenteric vessel vasculitis can cause bowel ischemia and may lead to mucosal ulcerations and possibly perforation, eventually requiring emergency laparotomy. Pancreatitis, necrotizing acalculous cholecystitis, and liver eosinophilic infiltrates have also been reported.
The kidney is not frequently involved in CSS. Overall, renal involvement occurs in less than one-quarter of CSS patients, far less common than in WG or MPA. The characteristic glomerular lesion of CSS is focal segmental glomerulonephritis with necrotizing features indistinguishable from the other ANCA-associated vasculitis; however, renal disease is considered milder and rarely causes renal failure. Investigators from the United Kingdom have reported higher frequencies of renal involvement in CSS (50%–80%) than other series, and in their experience renal outcome seemed similar to that of MPA. A possible explanation for this discrepancy is that their patients were all drawn from a kidney disease unit. Finally, other renal lesions have been described in CSS, including eosinophilic interstitial infiltrates and IgA nephropathy.
Other organ systems may also be affected in CSS. Central retinal artery and vein occlusion, salivary gland enlargement, myositis, and vasculitic involvement of the breast have been described. Patients with CSS also seem to have a propensity for thromboembolic events, similar to that seen in patients with WG.
Clinical Significance of ANCAs in CSS: Just One Disease Entity?
CSS has been traditionally included in ANCA-associated vasculitis together with WG and MPA. The prevalence of ANCAs in CSS is less consistent than in WG or MPA, reported in approximately 40% of cases. The ANCA immunofluorescence pattern is usually perinuclear with specificity for myeloperoxidase (MPO) by enzyme-linked immunosorbent assay. Only a minority of patients have cytoplasmic ANCAs with antibodies to proteinase 3. Recently, it has been shown that different clinical phenotypes could be observed according to the presence or absence of ANCA, thereby suggesting different disease pathogenetic mechanisms in CSS. In 2005, for the first time, 2 different studies, one from the French Vasculitis Study Group and the other from Italy, compared CSS patients’ demographic, clinical, and laboratory features according to their ANCA status at diagnosis and described different disease phenotypes according to the presence or absence of ANCAs. Previously, only Keogh and Specks, retrospectively studying 74 CSS patients, had found an increased prevalence of central nervous system involvement in ANCA-positive patients. In all other series, no correlation was detected between ANCAs and CSS clinical manifestations, perhaps due to the small number of patients enrolled.
Sablé-Fourtassou and colleagues with the French Vasculitis Study Group described the clinical and serologic features of 112 CSS patients. Forty-three of 112 patients (38%) were ANCA positive and their positive ANCA status was associated with renal involvement, peripheral neuropathy, and biopsy-proved vasculitis. In the same series, a negative ANCA status was associated with heart disease and fever. Similar results were also found by Sinico and colleagues and published in the same year. ANCAs were present in 35 of the 93 Italian patients (37.6%) included in the study. ANCA positivity was significantly associated with higher prevalence of renal disease and pulmonary hemorrhage and, to a lesser extent, with other organ system manifestations (purpura and mononeuritis multiplex) but with lower frequencies of lung disease and heart disease. ANCA-positive patients also tended to relapse more frequently. Furthermore, in both studies, vasculitis was documented less frequently in histologic specimens from ANCA-negative patients in comparison with ANCA-positive patients.
On the basis of these findings, it has been hypothesized that, as in WG and MPA, ANCAs might contribute directly to the endothelial vasculitic damage in the ANCA-positive subset, whereas in the ANCA-negative subgroup, eosinophils might be directly responsible for tissue damage by the release of harmful cationic proteins, such as the eosinophilic cationic protein (implicated in cardiotoxicity), and the eosinophil-derived neurotoxin. Overall, these data suggest that CSS, when presenting as a form of necrotizing small-vessel vasculitis, is a true ANCA-associated disease characterized by the positivity for MPO ANCAs. Alternatively, the ANCA-negative subset seems clinically characterized by eosinophil tissue infiltration resulting in fibrotic organ damage and seems to belong more to the spectrum of HESs rather than to the spectrum of ANCA-vasculitis. The hypothesis of the existence of 2 subsets in CSS is now indirectly supported by the findings of an increased frequency of HLA-DRB4 in CSS patients and its association with the ANCA-positive disease subset. Although the existence of 2 clinical phenotypes may reveal the existence of different pathogenetic pathways in CSS, it should be stressed, that in these studies, the ANCA-positive and ANCA-negative dichotomy is not absolute and overlap between the 2 phenotypes is common. Moreover, the therapeutic implications in distinguishing these 2 subsets are far from clarified, because ANCA had no obvious effect on remission or survival. Thus, it is premature to propose different therapeutic strategies, according to pathogenic subsets.
CSS Laboratory Features and the Lack of Disease Biomarkers
Apart from ANCAs, many other laboratory abnormalities have been described in CSS. Nonetheless, the vast majority of these laboratory features are largely nonspecific and include normochromic normocytic anemia and elevated erythrocyte sedimentation rate (ESR) or C-reactive protein, especially during the phase of active vasculitis. In addition, elevated serum IgE has been observed in up to 75% of patients and serum rheumatoid factor has been detected in as many as 60% of patients.
Alternatively, peripheral blood eosinophilia (>10% on differential white blood cell count or >1.5 × 10 9 /L) is probably the hallmark of CSS and may characterize any stage of the disease. At the onset of the disease, the absence of eosinophilia may be explained by prior corticosteroid treatment for asthma. There is sometimes a close association between the degree of eosinophilia and the activity of the vasculitic disease. A rise in the eosinophilic count may precede a relapse of the vasculitis; however, these 2 features may be dissociated and peripheral blood eosinophilia might not recur during relapse. There is also evidence that peripheral blood eosinophils count may correlate with serum eosinophil cationic protein. Overall, despite the fact that hypereosinophilia and eosinophil cationic protein may be hallmarks of CSS, their diagnostic value is limited because they do not distinguish CSS from the other HESs. Moreover, no laboratory tests serve as reliable, feasible, and standardized disease biomarkers that reflect disease activity over time and can guide therapeutic intervention.
CSS, HES, and Other Systemic ANCA-Associated Vasculitis: Elements of Differential Diagnosis
The differential diagnosis of CSS includes several different diseases, mainly HES and the other forms of ANCA-associated vasculitis, such as MPA and WG.
HES is chronic condition characterized by a persistent and sustained peripheral blood eosinophilia, exceeding 1500 cells/μL for more than 6 consecutive months, which is responsible for the development of organ dysfunction or damage. HES shares several clinical and histologic features with CSS. The target organs in the 2 syndromes are similar. The clinical picture of HES might be characterized by lung, cardiac, PNS, gastrointestinal, and skin involvement. The organ involvement and damage is directly related to eosinophilic infiltration; signs of vasculitis and ANCA are completely absent in HES. No clinical manifestations are considered pathognomonic of HES or CSS ( Table 1 ). Late-onset asthma, however, is rare in HES (even though bronchial hyperactivity may be present) whereas endomyocardial fibrosis is widely considered a more typical finding of HES rather than of CSS. Recently, the diagnosis of HES has been facilitated by molecular biology techniques because specific mutations have been identified in some subsets of this syndrome. The identification of FIP1-like 1 platelet-derived growth factor receptor-alpha in cases of HES or T-cell antigen receptor rearrangements can be used routinely to differentiate the 2 conditions. There is an area of uncertainty that includes patients with CSS without the histologic evidence of vasculitis or ANCA and HES with negative molecular genetic testing. In these cases, it may be difficult to distinguish CSS from HES. Additional research may make it easier to differentiate the two conditions.
|Clinical Manifestations||Hypereosinophilic Syndrome||Churg-Strauss Syndrome|
|Heart involvement||Endocardial fibrosis, restrictive cardiomyopathy||Pericardial involvement, myocardial involvement |
Small vessel vasculitis
|Asthma||Rare||Almost always present|
|Skin involvement||Urticaria||Purpura, urticaria|
|Biopsy proved vasculitis||Absent||Present (especially in late phase)|
|ANCAs||Absent||Present in approximately 40% patients|