Microscopic polyangiitis (MPA) is a rare antineutrophilic cytoplasmic antibody (ANCA)–associated systemic vasculitic disease that affects small-caliber blood vessels.1 The Chapel Hill Consensus Nomenclature of vasculitides describes MPA as pauci-immune disease with absence of granuloma formation or prominent eosinophilia associated with myeloperoxidase (MPO) ANCA.1–5 Although MPA is a multisystem disease that may affect the lungs, kidneys, nerve, and skin, the literature addressing disease manifestations is limited by the partitioning of case recognition and management among various subspecialties. The purpose of this retrospective cohort study was to assess the rate of pulmonary manifestations in a large series of patients with a diagnosis of MPA.
STUDY DESIGN AND PATIENT POPULATION
After local institutional review board approval was obtained, we conducted a single-institution, retrospective chart review to identify all patients diagnosed with MPA over the 10-year period from September 1, 2001, through September 1, 2011. Because MPA does not have a specific International Classification of Diseases, Ninth Revision (ICD-9) code, we reviewed electronic medical records of patients with related diagnostic codes for vasculitis, Wegener granulomatosis, and polyarteritis nodosa. In addition, we reviewed the surgical pathology department’s electronic database (SunquestPowerPath), and the immunology laboratory’s listings of all patients with a positive p-ANCA or MPO (Mayo Reference Lab). The combination of all searches identified 40 patients who met case definitions for inclusion based on either pathology meeting Chapel Hill Consensus Nomenclature description or combined clinical and serologic diagnosis.3 All patient charts included in the case series were reviewed by each author independently. The authors then collaborated to further review and ensure that inclusion criteria were met. Patients with questionable MPA diagnoses were excluded.
Multiple variables were collected from each chart and included age, sex, race, occupation, medications, smoking history, history of diabetes, hepatitis, hypertension, presenting symptoms, pulmonary and renal function studies, pathology and imaging reports, subsequent treatment, and the managing subspecialty. Patient survival was obtained from the electronic medical record and the Social Security Death Index (http://www.genealogybank.com).
Median age of the 40 patients was 61 years (range, 14–83 years). Sixty-two percent were female, 70% were white, and 20% were Hispanic (Table 1). Comorbidities included hypertension in 23 patients and diabetes mellitus in 8. Half of the patients had a history of smoking tobacco, either at the time of diagnosis or before diagnosis.
Onset of illness appeared to be within days to weeks of presentation. The most common presenting complaint was cough. At time of diagnosis, cough was present in 42.5% (n = 17), hemoptysis in 40% (n = 16), and dyspnea in 22.5% (n = 9) (Table 2).
At presentation, MPA was affecting the lungs and kidneys of 75% (n = 30). There was renal involvement without pulmonary manifestations in 12.5% (n = 5), and isolated pulmonary involvement was found in 2 patients (5%), one of whom had tracheal stenosis. Pulmonary hemorrhage occurred in 47% (n = 19). Twenty-five patients had a serum creatinine of 1.5 or greater at time of diagnosis, and 27.5% (n = 11) progressed to end-stage renal disease and hemodialysis. Peripheral neuropathy was noted in 2 patients.
Thirty-eight patients had positive p-ANCA. The 2 patients who were negative for p-ANCA had positive MPO. The 38 patients tested had a positive MPO. Antinuclear antibody titers were checked in 39 patients, and 35 were either negative or low titer. Four had a significant antinuclear antibody titer of more than 1:1280, but did not meet diagnostic criteria for a specific connective tissue disease.
During the course of illness, chest imaging showed pulmonary infiltrates in 92% (n = 37). All patients had plain chest films, in which the infiltrates were almost equally described as interstitial and alveolar, with many showing a mixed pattern. Twenty-seven patients had documented computed tomography (CT) scans, of which 2 were high-resolution CT scans. On CT scan, the majority of the abnormalities were ground-glass opacities and alveolar infiltrates, which varied in distribution. Only 2 patients were noted to have honeycombing, which was mild and peripheral only. Only 1 patient had a mass-like configuration, not clearly related to MPA; otherwise, there were no masses, and well-formed nodules were rarely seen. No cavities were noted. Pleural effusions were noted in 15% (n = 6), all of which were small and accompanied by interstitial and alveolar infiltrates.
Pulmonary function testing (PFT) at or after the time of diagnosis was conducted in 17 patients (42.5%). The majority (12 of those 17) had an FEV1/FVC (forced expiratory volume in 1 second/forced vital capacity) of 70 or greater (consistent with nonobstructive lung disease). Total lung capacity was measured in 6 patients, and 2 had values below the lower limit of normal with a median of 83% predicted of all measured (range, 51%–100% predicted). In the 13 patients who had a diffusion capacity of carbon monoxide (DLCO) measured, 12 were below the lower limits of normal, and the median was 56% predicted (range, 45%–81% predicted). Fifteen of the 17 patients with PFTs performed had abnormal chest radiographs. Of the 2 remaining patients, 1 had only spirometry performed, which was normal. The patient with tracheal stenosis had a normal chest radiograph, normal spirometry, and a mildly decreased DLCO of 66% predicted.
Bronchoscopy was performed in 60% (n = 24) (Table 3). Twelve had evidence of active alveolar hemorrhage, and 7 had evidence of recent hemorrhage. Open or video-assisted thoracic lung biopsy was performed in 12.5% (n = 5). Renal biopsy was performed in 60% (n = 24). Other tissue diagnoses were obtained from sural nerve in 2 patients, quadriceps muscle in 1, and aortic valve in 1. In 6 patients, no tissue diagnosis was obtained, and the patients were diagnosed based on clinical presentations and positive serology.
Treatment usually included corticosteroids and cytotoxic therapy, but medication choices, doses, and duration were quite variable. Of particular note, 26 patients received cyclophos phamide (in addition to corticosteroids) with median of 6.5 months of oral and 6 months of intravenous. Six patients were treated with corticosteroids alone; 3 patients underwent plasmapheresis, and 2 were treated with rituximab (Table 4). In our series, 22 were managed by nephrology, 9 by pulmonary, and 9 by rheumatology. However, the majority of patients were followed by multiple subspecialties
Using a final contact time of February 1, 2013, 14 patients (35%) have died since the beginning of the study period in 2001. One patient died directly of MPA, 1 died of a gastric ulcer on high-dose corticosteroids, 1 died of aspiration pneumonia, 1 died of cardiac valvular disease, and 1 died of acute myelogenous leukemia. The other 7 were lost to follow-up, with cause of death unknown. Mean follow-up time among patients who were survivors at the final contact time described above was 4.8 years, whereas mean follow-up of all patients, irrespective of survival, was 4.1 years. A Kaplan-Meier curve (Figure) was constructed using the date of diagnosis of 37 subjects. One was excluded secondary to uncertainty of date of diagnosis, and 2 were excluded because they were lost to follow-up.
FIGURE. Kaplan-Meier...Image Tools
The most important findings of our retrospective, cohort study of patients diagnosed with MPA are the recognition of an increased incidence of pulmonary manifestations compared with published literature and the occurrence in our Hispanic population.
Microscopic polyangiitis has a reported incidence of 2.7 to 94 per 1 million.6–8 Existing literature suggests a higher frequency in European populations, but populations in other parts of the world have not been well studied. Other studies indicate that the frequency does not appear to be affected by latitude.4,6–8 Our observation regarding the Hispanic population could be explained by demographics in our region, but the genetic predisposition for MPA recently described by Lyons et al9 has not been explored in this subgroup of patients. A slight female predominance in our series contrasts with previous literature.
Previous case series of MPA have been reported mainly in the nephrology literature and emphasize renal manifestations. In our review, we screened our subjects by ICD-9 codes, serological testing, and a surgical pathology database. Pulmonary involvement without renal involvement was found in 2 of the 40 subjects, suggesting that MPA can be present without evidence of renal manifestations.
Microscopic polyangiitis has been previously characterized by initial constitutional symptoms with the development of glomerulonephritis in the majority of patients and pulmonary involvement in only 22% to 55%.4,10 Agard et al11 found that patients presented most commonly with generalized symptoms, such as weight loss, fever, asthenia, myalgias, and arthralgias. Pulmonary capillaritis or diffuse alveolar hemorrhage has been seen in about 10% to 30% of patients, which constitutes a form of pulmonary-renal syndrome in the setting of glomerulonephritis.1,3,4,6,10 One smaller retrospective study of 29 cases of MPA focused on pulmonary complaints and found cough as the initial presenting symptom in 90% and hemoptysis in 79%.12 In our larger study, we found a similar incidence of pulmonary involvement, including 75% of patients with respiratory complaints at presentation and 92% with pulmonary infiltrates during the course of the disease.
Radiography has been useful in identifying pulmonary involvement in MPA. In one series, CT imaging demonstrated ground-glass opacities in 94%, but other studies have identified multiple other patterns.3,6,13 In our series, pulmonary infiltrates were reported in 92% of patients. Imaging was usually ordered to evaluate the symptoms, rather than to specifically characterize the disease. When pulmonary hemorrhage was bronchoscopically confirmed, infiltrates could be specifically ascribed to vasculitic changes. However, when an invasive procedure was not performed, our review would be unable to exclude the contribution of comorbidities such as congestive heart failure, renal failure, or infection. Pleural effusions were small and were likely associated with comorbidities, such as congestive heart failure and renal failure.
In previous series, the patterns observed on PFTs have been either restrictive or obstructive with a reduced DLCO.6 Our results demonstrated a primarily nonobstructive pattern and decreased DLCO in those tested, but nonpulmonologists underutilized the test in patients with pulmonary complaints. The utility of serial PFTs has not been studied.
Tissue pathology is important in diagnosing MPA. In our study, we found only 1 diagnostic transbronchial biopsy, which is in agreement with previous reports that bronchoscopy has a limited role in diagnosing MPA. However, bronchoscopy with inspection and bronchoalveolar lavage may be helpful in identifying pulmonary hemorrhage.12 Renal and surgical lung biopsies have a high yield and are extremely valuable in the diagnosis of glomerulonephritis and pulmonary vasculitis, respectively.4,6 Other biopsy sites can be utilized to demonstrate small vessel vasculitis in the appropriate clinical context, as we saw in 4 patients. With the use of serology, a clinical diagnosis may be acceptable when a patient is too ill for invasive procedures.
Differentiating between MPA and granulomatosis with polyangiitis (formerly Wegener granulomatosis) should not delay treatment when a pauci-immune small vessel vasculitis is identified.3,14 In granulomatosis with polyangiitis, cyclophosphamide and high-dose corticosteroids have been considered standard therapy and have been shown to markedly improve morbidity and mortality.3 Induction therapy with methylprednisolone and an additional immunosuppressant is important for rapid control of inflammation.14 Treatment of severe disease may include cyclophosphamide, corticosteroids, and plasmapheresis.5 Plasmapheresis has been shown to be beneficial for pulmonary hemorrhage and severe kidney disease.14 Recently, rituximab has been shown to be noninferior to cyclophosphamide for induction in severe ANCA-associated vasculitis.15
We found variability in the subspecialty initially managing the patients. Of the patients treated by the pulmonary service, about half were in the medical intensive care unit and were also followed by the nephrology service. In evaluating the types of treatment given to our patients, we found a large array of medication combinations (Table 4). Because of the multiple treatment combinations chosen and the various presentations leading to treatment, it would be difficult to ascribe specific benefit to 1 treatment versus another; however, cyclophosphamide was used in 26 (65%) of patients. Cyclophosphamide toxicity was not suspected in any of the patients with radiographic changes. Variability of subspecialty services providing long-term follow-up was observed as well, and only 6 patients were followed up by all 3 subspecialties.
We encountered some limitations in our retrospective review. First, a specific ICD-9 code for MPA does not exist, which posed difficulty in identifying subjects with a specific diagnosis of MPA. However, use of the p-ANCA and MPO data aided in the search for more patients with the diagnosis. Absence of the prospective methodology in our series prevented correlation of symptoms with imaging and other studies. Finally, given the nature of our retrospective review, not all the variables sought were obtainable because of incomplete information in the medical record and lack of uniformity in documentation. Our review failed to identify common inflammatory triggers for MPA.
In a large, single-center study of patients with MPA, we identified pulmonary involvement in 80% on the basis of symptoms alone and 92% when radiographic changes were included. This result significantly exceeds the previously reported 25% to 50% pulmonary involvement. In general, respiratory symptoms may be underreported by the evaluating physicians. When features in addition to pulmonary hemorrhage are recorded, MPA involved the pulmonary system in almost all of our patients.
Microscopic polyangiitis has a higher prevalence in the Hispanic population than has been determined in previous studies dominated by white populations. More studies may be needed to determine further genetic predispositions within the diverse Hispanic patient population.
There are no clear guidelines to direct evaluation and management of MPA patients. Consistent communication between pulmonary, nephrology, and rheumatology services could improve our understanding of the disease process as well as improve patient care.
The authors thank Sheree Thomas, BS, System Data Metrics, for assistance in subject identification; Alejandro Arroliga, MD, for reviewing the manuscript; Robert S. Beissner, MD, PhD, associate professor, Department of Pathology, Texas A&M University College of Medicine, College Station, for assistance with subject identification and pathology; and Juhee Song, PhD, for assistance with biostatistics analysis.
© 2014 by Lippincott Williams & Wilkins, Inc.