Antineutrophil Cytoplasmic Antibodies and the ChurgStrauss Syndrome

Context Patients with the ChurgStrauss syndrome, a rare small-sized vessel vasculitis, sometimes have antineutrophil cytoplasmic autoantibodies (ANCAs). Contribution In this cross-sectional study, 43 of 112 (38%) patients who recently had received a diagnosis of the ChurgStrauss syndrome were ANCA-positive. Compared with ANCA-negative patients, ANCA-positive patients more often had peripheral neuropathy (84% vs. 65%) and renal manifestations (35% vs. 4%). Also, ANCA-negative patients more often had fever (55% vs. 30%) and cardiac manifestations (49% vs. 12%). Implications Phenotypes of patients with ANCA-positive and ANCA-negative ChurgStrauss syndrome may differ. The Editors The ChurgStrauss syndrome is a rare, small-sized vessel vasculitis that develops in patients who have usually had asthma and other allergic manifestations for several years or months. The syndrome is characterized by pulmonary involvement and extrapulmonary manifestations that can be serious and life-threatening when the heart, central nervous system, gastrointestinal tract, or kidneys are affected. It is associated with antineutrophil cytoplasmic antibodies (ANCAs) in 39% to 59% of patients (1, 2). For most ANCA-positive patients, the fluorescence pattern is perinuclear and IgG recognizes myeloperoxidase, as assessed by enzyme-linked immunosorbent assay (ELISA). No analysis of a large homogeneous cohort of patients with the ChurgStrauss syndrome has been reported since systematic ANCA testing became available, and the association of ANCA status with clinical characteristics and outcome has not been evaluated. We aimed to describe the clinical characteristics of patients who recently received a diagnosis of the ChurgStrauss syndrome and who were included in prospective therapeutic trials organized by the French Vasculitis Study Group (FVSG) and to compare the characteristics of patients with and without ANCA. Methods Patients and Treatments All patients had ChurgStrauss syndrome that satisfied the classification criteria established by the American College of Rheumatology (3) or the Chapel Hill nomenclature (4). We selected the cohort of 112 patients from those who were recruited by members of the FVSG and the European Vasculitis Study Group (EUVAS) and were enrolled in different prospective trials. We initially screened the report forms of 127 patients. We did not select 15 patients for several reasons: No data were forwarded for 7 patients, diagnosis was not confirmed for 6 patients, and 2 patients declined to give written informed consent. We enrolled all consecutive patients with the ChurgStrauss syndrome who met the inclusion criteria, except patients who were older than 75 years of age. We recruited most patients in France (58 centers) and other patients in Belgium (3 centers), Latvia (1 center), and the United Kingdom (1 center). The trials aimed to evaluate the optimal treatment of the ChurgStrauss syndrome. We considered only patients who had recently received a diagnosis, were older than 18 years of age, and were presenting their first flare for the trials. Treatment choice did not consider the patient's ANCA status. Once we confirmed the diagnosis, we stratified patients according to the 5-factor score, which we prospectively devised and retrospectively validated to predict the overall mortality of patients with the ChurgStrauss syndrome according to the presence or absence of symptoms at diagnosis (5). The following items of poor prognosis make up the 5-factor score: classes of serum creatinine levels (140 mol/L [1.58 mg/dL] and >140 mol/L [>1.58 mg/dL]) and proteinuria (1 g/d and >1 g/d), severe gastrointestinal tract involvement, cardiomyopathy, or central nervous system involvement. We assigned 1 point for the presence of each factor. We also retrospectively applied the Birmingham Vasculitis Activity Score in its first presentation (6) at diagnosis. The score is a clinical index of disease activity based on symptoms and signs in 9 separate organ categories (systemic signs, skin, mucous membranes and eyes, earnosethroat, chest, heart and vessels, gastrointestinal, kidney, and central nervous system). According to the number and severity of clinical and biological symptoms, we accorded a predefined number of points to each category, then added the individual category scores to yield a maximum possible score of 63 points. We considered disease features only when they are attributable to active vasculitis. Because the Birmingham Vasculitis Activity Score was created after the therapeutic protocols had been designed, we did not use it to select the patient's treatment. We also did not use it for follow-up or to define remission, failure, or relapse. We used only clinical and biological manifestations for this purpose. We assigned patients with a 5-factor score of 1 point or greater to receive 6 or 12 pulses of cyclophosphamide in combination with corticosteroids. We assigned patients without a poor prognosis factor (5-factor score of 0 points) to steroids alone and to receive immunosuppressants only when steroid therapy failed. Because the therapeutic trials are ongoing, their results are not yet available. We prospectively recruited patients between 1995 and December 2002. Analysis Once the ChurgStrauss syndrome was diagnosed, we recorded all data in a computerized databank. The pathologist provided histologic reports and slides that could be reviewed if desired. We obtained clinical reports, recorded on standardized report forms and completed by the treating physician, every month during the first year of follow-up, then every 3 months for the following year. When disease-related events occurred during follow-up, we obtained additional reports. We tested serum specimens for the presence of ANCA by indirect immunofluorescence, according to EUVAS recommendations (7), by using serum diluted to 1:16 and ethanol-fixed neutrophils. When we detected ANCA, we recommended but did not require ELISA determination of specificity (antimyeloperoxidase or antiproteinase 3). We performed ANCA tests at diagnosis and before starting treatment. When ANCA test results were initially positive, we checked results at each visit as required by the therapeutic protocols. We routinely tested ANCA in the immunology department of each participating hospital by using commercially available tests. The reproducibility and specificity of biological and immunologic tests are guaranteed, in France, by a national commission of the Ministry of Health. We performed some routine biological analyses (complete blood counts, creatininemia, electrolyte levels, proteinuria, hematuria, and aminotransferase levels) and chest radiography at entry and at each visit as requested by the protocol. Electrocardiography was compulsory at entry but was optional during follow-up. We performed more specific investigations (echocardiography and catheterization) when indicated by clinical manifestations and requested by the treating physician. Some data are missing: Information on a triggering factor was not identified in 3 patients, no data were obtained on earnosethroat status in 4 of 112 patients, and creatinine clearance could not be calculated for 9 of 112 patients. In all but 4 patients, C-reactive protein or erythrocyte sedimentation rate was reported. For 1 of 71 patients with biopsy-confirmed diagnoses, the complete pathologist's report form was not sent, and we saw only the conclusion. Definitions We considered asthma to be severe when characterized by continuous dyspnea, when continuous administration of corticosteroids was required, or when oximetry and other pulmonary parameters necessitated hospitalization in an intensive care unit. We diagnosed specific cardiomyopathy after electrocardiography, echocardiography, and coronary artery arteriography in some patients to exclude another cause. Diagnosis was sometimes proven by biopsy. We considered histologically proven glomerulonephritis or proteinuria greater than 0.4 g/d and microscopic hematuria without urinary infection to be sufficient to diagnose renal involvement. We recorded clinical remission when clinical symptoms stabilized or regressed and initial laboratory abnormalities (except ANCA status) returned to normal. We defined relapses as the appearance of new systemic manifestations of vasculitis or worsening of at least 1 initial disease manifestation. Role of the Funding Sources The Ministre de la Recherche and the Institut National de la Recherche Mdicale, Paris, France, provided grants, which were used to monitor the therapeutic studies. The Hospices Civils de Lyon, Lyon, France, sponsored the therapeutic trials and was responsible for all of the legal aspects of the studies but did not support the studies financially. The principal investigators are employees of the Assistance PubliqueHpitaux de Paris. Members of the FVSG and the EUVAS included patients. The authors reviewed all of the data and wrote the manuscript. Results ANCA Status Testing We detected ANCA giving a perinuclear or cytoplasmic immunofluorescence-labeling pattern in serum specimens from 43 (38%) patients: 39 perinuclear ANCApositive patients and 4 cytoplasmic ANCApositive patients. Enzyme-linked immunosorbent assay showed that serum specimens of 34 of 39 patients with perinuclear ANCA (5 patients' specimens were not tested) had antimyeloperoxidase specificity, but myeloperoxidase or proteinase 3 activity was not found in the 4 patients with cytoplasmic ANCA. We did not test other antigens. Demographic Characteristics and Clinical Symptoms Tables 1 and 2 detail the clinical characteristics of the 112 patients (57 women and 55 men; mean age, 52 years [SD, 15]) who participated in the study. Every patient had asthma, which preceded the first manifestation of the ChurgStrauss syndrome in 101 patients. Asthma and the ChurgStrauss syndrome occurred simultaneously in 9 patients, whereas asthma occurred 9 and 17 months after other clinical manifesta